lapin-ci

APPROVED:
Karen Toussaint, Major Professor
Einar T. Ingvarrson, Committee Member
Manish Vaidya, Committee Member
Richard Smith, Chair of the Department of
Behavior Analysis
Tom Evenson, Dean of the College of Public
Affairs and Community Service
Mark Wardell, Dean of the Toulouse Graduate
School
THE MEASUREMENT AND ENHANCEMENT OF RAPPORT BETWEEN
BEHAVIORAL THERAPISTS AND CHILDREN WITH AUTISM
Carly Ilyse Lapin, B.S.
Thesis Prepared for the Degree of
MASTER OF SCIENCE
UNIVERSITY OF NORTH TEXAS
December 2014

Lapin, Carly Ilyse. The Measurement and Enhancement of Rapport Between Behavioral
Therapists and Children with Autism. Master of Science (Behavior Analysis), December 2014,
72 pp., 7 tables, 22 figures, references, 32 titles.
Rapport has been acknowledged as an important variable in therapeutic contexts. The
current evaluation defined and assessed rapport quality between children with autism and
behavioral therapists based on behavioral correlates. In addition, the author evaluated the effects
of an operant discrimination training procedure to enhance rapport levels for therapists with low
levels of rapport. More specifically, the current study evaluated: (a) if the discrimination training
procedure would establish therapists’ social interactions as a discriminative stimulus and (b) if
social interaction would function as a conditioned reinforcer for novel responses. Results suggest
that the discrimination training procedure was successful in conditioning social interaction as a
reinforcer for all child participants, and as a result, rapport increased.

Copyright 2014
by
Carly Ilyse Lapin
ii

ACKNOWLEDGMENTS
I would like to thank Dr. Karen Toussaint for her guidance and mentorship throughout
graduate school. She has provided me with so many learning opportunities that have facilitated
my growth and competency as a behavior analyst. During this thesis, she has given me guidance
but also freedom to explore my own ideas, and for that I shall always be thankful. Dr. Toussaint
has become a role model to me, and I hope to continue making her proud in the upcoming steps
of my career as a practitioner.
I would also like to thank Dr. Einar Ingvarsson who has become another important
mentor during this thesis. He has been instrumental in the formation of this thesis topic. I would
also like to thank John Carter and Kyle Wiggly who spent a great deal of time coding videos and
collecting IOA. I would like to thank Manish Vaidya for taking time out of his busy schedule to
sit on my committee. Finally, I would like to thank my family for their support, encouragement,
and most importantly, patience. Given the time consuming nature of a thesis, they have been
patient with my time allocation and provided with me unconditional love and courage throughout
my graduate school experiences.

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS .............................................................................................................iii
LIST OF TABLES.......................................................................................................................... v
LIST OF FIGURES ....................................................................................................................... vi
CHAPTER 1 INTRODUCTION.................................................................................................... 1
CHAPTER 2 GENERAL METHOD............................................................................................ 11
CHAPTER 3 INITIAL ASSESSMENTS..................................................................................... 15
CHAPTER 4 METHOD ............................................................................................................... 18
CHAPTER 5 PROCEDURE......................................................................................................... 28
CHAPTER 6 RESULTS............................................................................................................... 32
CHAPTER 7 GENERAL DISCUSSION..................................................................................... 38
APPENDIX A TREATMENT INTEGRITY CHECKLIST FOR RESPONSE ASSESSMENT 56
APPENDIX B TREATMENT INTEGRITY CHECKLIST FOR BASELINE & POST-
TRAINING SESSIONS................................................................................................................ 58
APPENDIX C TREATMENT INTEGRITY CHECKLIST FOR DISCRIMINATION
TRAINING ................................................................................................................................... 60
APPENDIX D SOCIAL VALIDITY QUESTIONNAIRE.......................................................... 63
APPENDIX E SOCIAL VALIDITY RESULTS ......................................................................... 65
REFERENCES ............................................................................................................................. 69
iv

LIST OF TABLES
Page
1. Formation of Dyads for Cole.............................................................................................43
2. Formation of Dyads for Zane.............................................................................................43
3. Formation of Dyads for Tommy........................................................................................44
4. Results of Response Assessment .......................................................................................44
5. Interobserver Agreement Results for Cole ........................................................................45
6. Interobserver Agreement Results for Zane........................................................................46
7. Interobserver Agreement Results for Tommy ...................................................................47
v

LIST OF FIGURES
Page
1. Baseline and post-training discrim. training......................................................................48
2. Discrimination training for Cole........................................................................................49
3. Discrimination training for Zane .......................................................................................49
4. Discrimination training for Tommy...................................................................................49
5. Child emitted rapport behaviors.........................................................................................50
6. Mutual rapport behaviors...................................................................................................51
7. Therapist emitted rapport behaviors ..................................................................................52
8. Cole’s approaches ..............................................................................................................53
9. Cole’s eye contact..............................................................................................................53
10. Cole’s body orientation......................................................................................................53
11. Cole’s physical contact ......................................................................................................53
12. Cole’s smiles......................................................................................................................53
13. Zane’s approaches..............................................................................................................54
14. Zane’s eye contact..............................................................................................................54
15. Zane’s body orientation .....................................................................................................54
16. Zane’s physical contact......................................................................................................54
17. Zane’s smiles .....................................................................................................................54
18. Tommy’s approaches.........................................................................................................55
19. Tommy’s eye contact.........................................................................................................55
20. Tommy’s body orientation.................................................................................................55
21. Tommy’s physical contact.................................................................................................55
22. Tommy’s smiles.................................................................................................................55
vi

CHAPTER 1
INTRODUCTION
Rapport (i.e., relationship quality) between therapists and children with autism is often
acknowledged as an important variable in the therapeutic context (Egan, 1975; Taylor & Fisher,
2010). McLaughlin and Carr (2005) conceptualized rapport as a setting event and demonstrated
that the likelihood of escape-motivated problem behavior was reduced when instructions were
delivered by staff members that self-identified as having pleasant social interactions with a client
or “good rapport.” The results suggest the importance of rapport quality as a contextual variable
that may influence the relationship between instructions and compliance.
In addition to facilitating learning opportunities, rapport per se is often a meaningful goal
for individuals with autism. Given that children with autism have characteristic deficits in social
interactions, reciprocal engagement in positive social interactions is often selected for
improvement (Strain & Shores, 1977). However, most behavioral interventions have focused on
improving the interaction between children and their peers; the social relationship between the
therapist and the learner has received substantially less attention in the literature (McConnell,
2002; White, Keonig, Scahill, 2007).
Given the importance of rapport, therapists are encouraged to “establish or build rapport”
before even beginning a treatment program for a child with autism (Taylor & Fisher, 2010).
However, this recommendation may be difficult to implement as rapport is generally not defined
in precise, measurable properties that would allow an individual to determine if rapport is
established. Rather, conceptual definitions are often provided to describe rapport with terms
such as “likeability” (Aronson, 1984), “empathy” (Roberts & Bouchard, 1989), or “mutual
understanding” (O’Toole, 2012) with convergence around the notion of a pleasant social
1

relationship. As a result, research aimed at identifying rapport between individuals with autism
and their caregivers has primarily focused on the use of specific rating scales that are subjective
in nature (Dunlap, Eno-Hieneman, Clarke, & Childs, 1995; Dunlap & Koegel, 1980; Koegel,
Dyer, & Bell, 1987). An operational definition of rapport would provide greater precision and
specificity.
Previous researchers have proposed an extended conceptualization of rapport that
includes identification of specific nonverbal correlates that comprise rapport. Tickle-Degnen and
Rosenthal (1990) proposed that rapport is a complex interaction which involves three
interrelating components: (1) mutual attentiveness, (2) positivity, and (3) coordination, and that
each of these components may be comprised of discrete behaviors. The first component of
rapport is mutual attentiveness, described as focusing and attending to the behavior of one
another. Tickle-Degnen and Rosenthal (1987) suggest that displaying mutual attentiveness can
be operationalized by measuring behaviors such as spatial configurations and bodily postures.
This suggests that approaches to another person, proximity to others (i.e., spatial configurations),
and body orientation (i.e., bodily postures) may be important behaviors indicative of rapport.
The second essential component is positivity, conceptually defined by a feeling of mutual
friendliness, warmth, and caring during an interaction. Accordingly, positivity may be
characterized by “the movement of the mouth, eyes, the brow, and the head in concert with one
another” (p. 289). This suggests that behaviors such as smiles and eye contact during dyadic
interactions may be important variables as well as physical contact. The final component of
rapport is coordination between a dyad. Coordination is defined in terms such as “balance,”
“harmony” and feelings of being “in sync” with one another (p. 286) and may be observed as
postural mirroring and interactional synchrony. Coordination highlights the nature of rapport as a
2

mutual process involving both members of a dyad. Together, these components suggest that
rapport is complex and the nonverbal behaviors of each member of the relationship could be
measured.
Gillis, Bernieri, and Wooten (1995) utilized the conceptualization offered by Tickle-
Degnen and Rosenthal (1990) and found that independent observers could reliably identify the
level of rapport between a dyad by measuring behavioral correlates. Gillis et al. (1995)
videotaped interactions of several dyads comprised of high school and undergraduate students
discussing a selected controversial topic (e.g., gun rights, abortion, etc.). Following the
discussion, dyad members rated their level of rapport using a self-report questionnaire that
contained 29 eight-point Likert scales. Next, the investigators reviewed brief video footage (e.g.,
50 seconds) of the dyadic interactions and created a graphical representation displaying the
percentages of several behavioral correlates of rapport that occurred during the interactions.
Further, Gillis et al. (1995) asked independent observers to determine the level of rapport based
upon these observable behaviors. For example, behaviors included adaptors (e.g., nervous
behaviors such as scratching, itching, etc.,), facial expressivity, mutual eye contact, gestures,
mutual silence, orientation, proximity, racial match, regulators, head nods, smiling, and postural
mirroring (see Gillis, Bernieri, & Wooten, 1995). Independent observers’ ratings were compared
to dyad members’ self-reported levels of rapport, and the results suggested correspondence
between the two measures.
The results of Gillis et al. (1995) demonstrate that independent observers’ objective
assessment of rapport matches self-reported, subjective ratings of rapport. However, previous
research, including that of Gillis et al. (1995), have typically focused on measuring rapport
between typically-developing dyads in prearranged contexts and situations (Babad, Bernieri, &
3

Rosenthal, 1987, 1989; Bernieri, Resnick, & Rosenthal, 1988). Measuring rapport between an
individual with autism and a caregiver during naturalistic interactions (i.e., playing) has received
considerably less attention.
One possible reason for the lack of research on measuring rapport within the field of
autism and related disorders is that individuals with autism may have deficits in vocal
communication. If an individual has little to no vocal language, it may present a challenge to
identify a child’s preference, level of rapport, and/or “likeability” for certain caregivers. For
example, in the study conducted by Gillis et al. (1995), each member of the dyad was able to
provide a lengthy and detailed self-report on the level of rapport they felt during an interaction.
This self-report was then compared with behavioral correlates coded by independent observers
who observed the interaction. Children with autism, especially those with deficits in vocal
behavior, may not be able to vocally describe the quality of their relationship with staff.
An alternative to reliance of vocal self-report is to measure behaviors indicative of
rapport (i.e., mutual attentiveness, positivity, and coordination). These behaviors may be
differentially observed in the presence of preferred therapists but not in the presence of novel or
less preferred therapists. In other words, it may be possible that a child will display higher levels
of approaches, smiles, eye contact, initiated physical contact, etc., in the presence of therapists
with whom they have established rapport with compared to therapists in which they have not
established rapport with.
McLaughlin and Carr (2005) provided a method to identify varying levels of rapport
between individuals with disabilities and various staff using a combination of rating scales, staff
interviews, and direct observation. Good rapport dyads met the following criteria: (a) a staff
member provided a high self-rating on the rapport Likert scale (Dunlap et al., 1995), indicating
4

they felt highly satisfied with their relationship quality with a given individual, (b) other staff
members ranked the dyad as demonstrating a high–quality relationship and (c) through direct
observations, the staff member was consistently selected by the individual with disability during
a preference assessment. The same measures were used to form the poor rapport dyads, in that
selection was based on a low self-rating made by the staff, a low rating made by other staff
during formal interviews, and a low rate of selections by the individual with disabilities (i.e.,
chosen rarely).
Although McLaughlin and Carr (2005) utilized both subjective (e.g., self-ratings made by
staff) and objective (e.g., preference assessments made by the participant) methods to describe
the relationship quality between an individual with disabilities and a caretaker, the field has yet
to define rapport in terms of observable and measurable responses. One purpose of the current
study was to increase the specificity of the current measures of rapport by measuring behavioral
correlates. Given the importance of establishing rapport between a therapist and a child with
autism, the second purpose of the current study is to enhance rapport of child-therapist dyads as
measured by increases in behavior correlates.
McLaughlin and Carr (2005) implemented a multicomponent intervention package with
“poor rapport dyads” that involved the use of responsivity training that focused on recognizing
and reinforcing participants’ communicative requests, strengthening turn-taking behaviors in the
context of mutually preferred activities, and the use of noncontingent reinforcement. During the
noncontingent reinforcement component, staff members were instructed to deliver reinforcers on
a time-based schedule (that is, noncontingently). This procedure is based upon seminal studies
on conditioned reinforcement in which pairing a previously neutral stimulus with an
5

unconditioned stimulus results in the neutral stimulus acquiring reinforcing properties and
becoming a conditioned stimulus (Fantino, 1977).
Several instructional manuals for professionals that provide behavioral treatment for
individuals with autism recommend the use of traditional pairing procedures to increase rapport
(Carr, McConnachie, Levin, & Kemp, 1993; Leaf & McEachin, 1999; Maurice, Green, & Luce,
1996). For example, Carr et al. (1993) recommend that rapport building can be achieved by (1)
stimulus-stimulus pairings, (2) stimulus-response pairings, and (3) mand training. To further
elaborate, Carr et al. (1993) suggest that the first step in rapport building is for instructors to pair
their social interaction with a learner’s reinforcers by delivering reinforcement without requiring
any response requirement from the learner. After about 2-3 days of “free pairings,” the second
suggested step is to require to the learner to approach the instructor in order to receive
reinforcement (i.e., stimulus-response pairings). After approach behavior has been established,
the last step is to require the learner to approach the instructor, but also engage in some type of
communicative act (e.g., vocalizing, pointing, reaching, etc.,) in order to request for the
reinforcer. It is suggested that each day, the learner and instructor set aside time to work on the
above steps in the context of playing games, sharing similar interests, and enjoying each other’s
company.
Carr et al. (1993) present general guidelines for ways in which an instructor can become
more “likeable” (p. 248). The article continues to emphasize that these steps may establish the
instructor as a discriminative stimulus for approach and communicative behavior and state that
the guidelines relate to establishing oneself as a generalized reinforcer.
This is similar to the suggestions provided by Leaf and McEachin (1999) who instruct
practitioners and parents that, “even if your child does not like social reinforcers such as smiles
6

and praise, by associating them with primary reinforcers (e.g., food, drink, favorite toy, etc.),
they will eventually become reinforcing as well” (p. 30). This statement suggests that pairing
non-preferred stimuli with a child’s preferred stimuli is an effective method in conditioning
reinforcers for children with autism. The suggestions presented by Leaf and McEachin (1999)
and Carr et al. (1993) are helpful strategies, but the articles fail to provide empirical evidence as
to the effectiveness of different methods (e.g., stimulus-stimulus, stimulus-response,
discrimination training) in establishing social stimuli as conditioned reinforcement for children
with autism.
In contrast to the traditional pairing account of conditioned reinforcement, is the
discriminative stimulus account. (Kelleher & Gollub, 1962). In operant discrimination training,
a neutral stimulus is first established as a discriminative stimulus (SD
) for a specific response. In
other words, reinforcement is delivered for a specified target response occurring in the presence
of the SD
but not in its absence. When differential responding occurs (i.e., target responses occur
only in the presence of the SD
), it suggests that the neutral stimulus (e.g., social interaction) has
acquired discriminative properties. In addition, it has been suggested that discriminative stimuli
will then acquire reinforcing properties (Holth, Vandbakk, Finstad, Grønnerud, & Mari, 2009;
Lovaas, Freitag, Kinder, Rubenstein, Schaeffer, & Simmons, 1966). However, research provides
inconclusive results as to which account best explains the behavioral mechanism responsible for
conditioned reinforcement effects or which account provides a more effective set of procedures
to condition novel stimuli as reinforcers (Williams, 1994).
Holth et al. (2009) compared a discrimination training procedure and a stimulus-stimulus
pairing procedure and evaluated the effectiveness of each on establishing various neutral stimuli
as reinforcers for children of typical and atypical development. Various auditory stimuli (e.g.,
7

“yay” sound from computer, door bell, cell phone ring, etc.,) and visual stimuli (e.g., smiley face
on computer monitor, yellow ball on a stick, a blue card on a stick, etc.,) were first identified as
neutral stimuli for participants; that is, contingent delivery of these neutral stimuli did not
increase response rates for participants.
After identification of neutral stimuli, participants experienced both the discrimination
training procedure and the classical conditioning (i.e., stimulus-stimulus pairings) procedure. The
experimenters differentially reinforced a response (the child’s behavior of taking a reinforcer off
the table and consequently consuming it) in the presence of a neutral stimulus during the
discrimination training procedures. In the absence of the neutral stimulus, child responses were
blocked and preferred edible and leisure items were not accessed. In the classical conditioning or
stimulus-stimulus pairing procedures, the neutral stimulus was presented for a brief period of
time immediately before the delivery of the reinforcer. The number of pairings remained the
same in both procedures. The results of Holth et al. (2009) indicate that discrimination training
was a more effective method in establishing neutral stimuli as conditioned reinforcers for 5 of
the 7 participants. More specifically, post-intervention measures demonstrated that the
participants emitted more arbitrary responses when the stimuli associated with the discriminative
stimulus procedure were delivered in comparison to contingent delivery of stimuli associated
with the pairing procedure.
Despite the documented success of the discrimination procedure, some limitations are
worth noting. First, the stimuli used (e.g., “yay” from a computer, a ball on a stick, etc.,) are not
necessarily “social stimuli” as they do not require an additional individual to be present in the
environment during the delivery of the stimulus. Second, the neutral stimuli assessment assigned
to the discriminative-stimulus procedures were associated with higher response rates relative to
8

those neutral stimuli assigned to the classical conditioning procedures, prior to any conditioning.
Although data during post-intervention demonstrate a significant increase in the rate of
responding following discrimination training, the integrity of the results is weakened given the
bias towards the stimuli used during discriminative stimulus procedures.
Given the limitations described above, the documented effectiveness of a discriminative
stimulus procedure on conditioning neutral social stimuli as reinforcers (specifically for children
with autism) remains limited. Therefore, the current study further evaluates the effects of a
discriminative stimulus procedure in conditioning neutral stimuli as reinforcers for children with
autism. It is presumed here that if a therapist has deficient rapport with a child and their social
interaction is found to be a neutral stimulus (i.e., does not increase responding when delivered
contingent upon a response), “pairing” via a discrimination training procedure may be an
effective method in establishing social interaction as a reinforcer.
The discriminative stimulus procedure differs from other conditioning procedures
including stimulus-stimulus and response-contingent pairings, which only involve the
presentation of SD
trials. Discriminative stimulus procedures may be a more optimal method for
pairing in that it requires observing or attending responding from the learner. For example,
during discrimination training, it requires the learner to engage in an observing response to the
stimuli (both SD
and S-delta) in order for differential responding to occur. In stimulus-stimulus
response pairings, the stimulus is presented immediately before (or with some overlap) to the
presentation of reinforcement, and often times, it does not require the learner to make any
contact with the stimulus in order to receive reinforcement. Therefore, interspersing SD
and S-
delta trials may be advantageous to other pairing procedures in that it may enhance the effects of
pairing (see Dinsmoor, 1995a, 1995b). The current study evaluates the effects of a discriminative
9

stimulus procedure on establishing a therapist’s social interaction (i.e., neutral stimulus) as a
discriminative stimulus and furthermore, a conditioned reinforcer for children with autism.
To summarize, the current study evaluated the following regarding rapport: (1) if children
with autism would exhibit differential levels of behavior correlates (e.g., eye contact,
approaches, etc.,) in the presence of a high-rapport therapist compared to a lack-of-rapport
therapist, (2) if therapists’ social interaction could be established as a discriminative stimulus
using a discriminative stimulus conditioning procedure, (3) if social interaction was successfully
established as a discriminative stimulus, it was further evaluated if social interaction would then
acquire reinforcing properties, and finally (4) if acquiring reinforcing properties (as a result of
the discriminative stimulus procedure) would enhance behavioral correlates of rapport.
10

CHAPTER 2
GENERAL METHOD
Selection of Child-Participants
Three children were recruited via flyers distributed to parents of children who attended a
university-based autism clinic in Denton, Texas. Participation criteria included having received a
diagnosis of pervasive developmental disorder in accordance with the Diagnostic and Statistical
Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria (American Psychiatric Manual,
2000) and currently receiving a minimum of eight hours per week of therapeutic services in
order to ensure experimenter/client availability. There was no selection criteria based on the
gender, age, or skill level of the child.
Cole is 8-year-old male diagnosed with autistic disorder and received approximately
thirty-five hours per week (7 hours per day) of 1:1 behavioral services. Cole communicated
using four-to-five word sentences and engaged in a high rate of vocal stereotypy (e.g.,
perseverative speech, echolalia) and motor stereotypy (e.g., skin picking, nail biting). His social
interactions typically involved parallel play with preferred leisure items that remained in his
individual treatment room. Transitioning Cole out of his treatment room typically presented
problem behavior, such as eloping, crying, and aggression (e.g., biting therapist attempts to block
elopement). In addition, he displayed preference for certain staff. For example, Cole would
repeatedly request for his preferred staff (e.g., “I want Rick back”) and repeatedly request for
novel staff to leave the environment (e.g., ”goodbye [name],” “please scoot back,” “please
leave”).
Zane is a 6-year-old male diagnosed with autistic disorder and received approximately
sixteen hours per week (8 hours, 2 times a week) of 1:1 behavioral services. Zane spoke in four-
11

to-five word sentences. Zane displayed severe problem behavior in the form of aggression (e.g.,
hitting, kicking), property destruction (e.g., breaking items, throwing items), stripping, public
urination, and self-injury (e.g., head banging, chin hitting). Zane’s social interaction with
therapists typically involved reciprocal play in the gym and motor lab. Examples include playing
tag, riding scooters, crawling through tunnels, and engaging in other age appropriate activities
(e.g., building blocks, playing Wii). Prior to the study, Zane did not vocalize a preference for one
therapist over another, and he appeared to enjoy the initial novelty of new individuals in his
environment.
Tommy is a 2-year-old male diagnosed with autistic disorder and received approximately
twenty hours per week (4 hours per day, 5 times a week) of 1:1 behavioral services. Tommy
spoke using one-to-two word sentences. Tommy’s social interactions typically involved playing
with toy cars and musical toys in the classroom, riding tricycles in the gym, and being pushed on
the swing in the motor lab. Prior to the study, Tommy did not vocalize preference for a therapist,
but informal observations (e.g., direct observation of Tommy with novel therapists) revealed that
he displayed avoidance behavior (e.g., crying, whining, running away) when novel therapists
attempted to interact with him.
Selection of Behavioral Therapist
Six behavioral therapists were recruited after meeting the selection criteria. Selection of
behavioral therapist was based on two criteria: (a) self-ratings made by behavioral therapists and
(b) selections by the child-participant. In order to be considered for participation in the lack-of-
rapport dyad, a behavioral therapist must have rated their relationship quality with a child-
participant as 0 to 3 on a 6-point Likert-type scale that assessed rapport (see McLaughlin & Carr,
2005; Dunlap et al., 1995). According to Dunlap et al. (1995), a rating of 0 to 3 indicated that the
12

behavioral therapist felt highly unsatisfied or neutral with their relationship with a given child-
participant.
In order to be considered for participation in the high-rapport dyad, a behavioral therapist
must have rated their relationship quality with a child-participant as a 4 or a 5 on the rapport
Likert scale, indicating they felt a high degree of rapport (i.e., highly satisfied) with their
relationship with a child-participant. The rapport Likert-type scale utilized in the present study
was adapted from McLaughlin and Carr (2005) and created by Dunlap et al. (1995).
Final inclusionary criterion in either the lack-of-rapport dyad or high-rapport dyad
involved assessing child-participants’ preference for behavioral therapists. Four behavioral
therapists, two with high self-ratings and two with low self- ratings on the rapport Likert-type
scale participated in the staff preference assessments which utilized a paired-choice assessment
method (Fisher, Piazza, Bowman, Hagopian, Owens, & Slevin, 1992). All participating
therapists briefly interacted with the child-participant (i.e., 30-seconds) in a random order prior
to the assessment. Subsequently, pairs of therapists stood an equal distance of approximately 4 m
in front of the child participant and approximately 1.5 m from each other. The experimenter
stood behind the child participant and instructed the child to “pick who you want to play with.”
Selections were indicated either vocally (i.e., stating the name of the therapist), or through
actions (i.e., approaching a therapist, pointing to therapist). Contingent upon selection of an adult
therapist, the child and chosen therapist interacted with each other for 30 seconds. The next
choice trial was presented to the child until all therapists were presented with all other therapists
in the four-person grouping. Therapist-participants selected the least number of times (e.g., never
or once) were considered as low-preferred therapists whereas those selected most frequently
were considered high-preferred therapists.
13

A second observer recorded interobserver agreement (IOA) on a trial-by-trial basis for
100% of the preference assessments. An agreement was scored if both observers recorded the
same selected stimuli (i.e., therapist) or both observed a no-response during the paired choice
trial. IOA was calculated by dividing the number of agreements by the number of agreements
plus disagreements and multiplied by 100. IOA was 100%.
Tables 1, 2, and 3 provide the results and summary of the formation of dyads. To sum,
lack-of-rapport dyads were formed between a child and adult participant if a staff member was
identified as a low-preferred therapist based upon both child selections and low self-ratings. In
contrast, high-rapport dyads were formed between a child and adult participant if a staff member
was identified as a high-preferred therapist based upon both child selections and high self-
ratings.
14

CHAPTER 3
INITIAL ASSESSMENTS
Reinforcer Selection
Prior to the start of the study, direct observation and informal interviews with the child’s
parents and/or caretakers were conducted in order to obtain information about the child-
participants’ preferred edibles or leisure items. A paired-choice preference assessment (Fisher et
al., 1992) was conducted in order to identify the top five preferred edible and/or leisure items for
each child.
The results of Cole’s preference assessment identified five preferred stimuli (in order of
rank) including Brach’s Fruitios® all natural fruit candy, a stuffed animal (e.g., black bear), a
sensory finger former toy, an iPad® mobile digital device, and a Ninja action figure. For Zane,
only edible items were used. The results of Zane’s preference assessment identified five
preferred stimuli (in order of rank) including bite size portions of Pepperidge Farm Goldfish®
snacks, Twix® caramel cookie bars, M&M’s® milk chocolate, Kit Kat® minis, and Haribo
Gummi Bears ®. For Tommy, high-preferred items included (in order of rank) bite size portions
of Kit Kat® M&M’s®, Goldfish®, Gummi Bears ®, and Twix®. The highly preferred (HP)
stimuli were used during a later phase in the Study.
A second observer recorded interobserver agreement (IOA) on a trial-by-trial basis for
each administered preference assessment. An agreement was scored if both observers recorded
the same selected stimulus during a paired-choice trial. IOA was calculated by dividing the
number of agreements by the number of agreements plus disagreements and multiplied by 100.
IOA was 100% across all child participants.
15

Response selection
Prior to start of the study, several discrete arbitrary responses (e.g., touching an index
card, moving a star across a line) were tested for properties of automatic reinforcement. That is,
the current study evaluated if child-participants would continue to engage in the arbitrary
response in the absence of programmed consequences. The experimenter physically prompted
(i.e., hand-over-hand) the first two responses but delivered no programmed consequences to
expose the participants to the contingency. Following two prompted responses, data was
collected on the rate of responses during a three-minute observation period. The response
materials were always present and in close proximity to the child but no further instructions or
programmed feedback were delivered.
Table 4 provides the results of the response assessment, an operational definition of the
two target responses, and the materials required to emit the response. The response was
considered eligible for use in a later phase of the study if less than three independent
(unprompted) responses occurred per session (i.e., rate of 1.0 responses per minute) for three
consecutive sessions. If a high rate of responding occurred, defined by four or more unprompted
responses in any given session (i.e., rate of 1.33 responses per minute), the response was
discarded from further use in the study.
For each participant, two responses were identified that were found to have no
automatically reinforcing properties and one was randomly selected as the target response to be
used during baseline and post-training conditions. The target responses are referred to as
response “A” for each child-participant (e.g., Cole response 1A, Zane response 2A, and Tommy
response 3A). A second response was also recorded, referred as response “B”, and the response
was probed throughout baseline and post-training conditions as a secondary measure.
16

A second observer recorded data in vivo for a minimum of 33% of response assessment
sessions for Cole, Zane, and Tommy. IOA was calculated using event recording by dividing the
smaller number of observed occurrences by the larger number of observed occurrences and
multiplied by 100 to yield a percentage. An occurrence was scored if an independent response
was observed. Mean agreement was 100% for all participants.
In addition to IOA, treatment integrity data was collected. Two independent observers
calculated treatment integrity during 66% of response assessments. In other words, 2 of the 3
response assessment sessions per child were scored for treatment integrity. One observer
calculated integrity in vivo, and another observer calculated integrity of the same sessions from
video footage (i.e., IOA of procedural integrity was taken). A checklist of four necessary steps
was used to score and calculate treatment integrity of the experimenter’s behavior during the
response assessments (see Appendix A). The checklist involved the following steps: (1)
Experimenter physically prompts the first response (2) experimenter physically prompts the
target response again, (3) the therapist did not deliver any interaction, feedback, or instructions
throughout the session, and (4) response materials remained in close proximity to the child at all
times. Percentage of treatment integrity was calculated by dividing the number of correctly
implemented steps by the total possible number of steps and multiplied by 100. Treatment
integrity was scored as 100% by both of the independent observers for all sessions during the
response assessment conditions.
17

CHAPTER 4
METHOD
Participants
Three lack-of-rapport dyads and three high-rapport dyads participated after meeting
selection criterion (see Tables 1, 2, and 3). The lack-of-rapport child-therapist dyads included
Cole & Jan, Zane & Raquel, and Tommy & Katie. All three therapists were perceived to have a
lack of rapport with a given child-participant based on the selection criteria used in the current
study (e.g., self-ratings made by the therapist and self-ratings made by the child participant
during preference assessments).
Jan was female and an undergraduate student who worked as a behavioral therapist. She
occasionally (e.g., approximately two-to-three times a month) provided therapeutic services to
Cole, as she was an identified substitute therapist for Cole when his regular staff members were
unavailable. Jan had a self-rating of 1 on the rapport Likert-type scale. Cole never selected Jan
during the paired-choice staff preference assessment.
Raquel was female and a graduate student who functioned as a supervisor of behavioral
therapists for Zane’s therapeutic services. She also occasionally (e.g., approximately four-to-five
times a month) served as a substitute therapist for Zane when his regular staff members were
unavailable. Raquel reported that rapport between the dyad was neutral (i.e., ranking of 3 on the
rapport Likert-type scale). Zane never selected Raquel during the paired-choice staff preference
assessment.
Katie was female and functioned as the case manager for Tommy. Katie reported that her
relationship quality with Tommy was neutral (i.e., self-rating of 3 on rapport scale). In addition,
Katie reported observing Tommy for approximately two hours a week and that her interactions
18

with him during these observation periods were minimal. Tommy never selected during the
paired-choice staff preference assessment.
In addition, three behavioral therapists (Rick, Ry, & Marcy) were found to have a
perceived level of high-rapport with a given child participant based on the selection criteria used
in the current study. The high-rapport dyads included Cole & Rick, Zane & Ry, and Tommy &
Marcy.
Rick was an undergraduate student in behavior analysis and was one of Cole’s primary
behavioral therapists at the time of the study. Rick had a self-rating of 5 on the rapport Likert-
type scale, indicating he felt a high degree of rapport and satisfaction during his interactions with
Cole. In addition, he was chosen 100% of the time during Cole’s paired choice preference
assessment (see Table 1).
Ry was one of the two primary behavioral therapists for Zane at the time of the study. Ry
had a self-rating of 5 on the rapport Likert scale and was chosen 100% of the time by Zane
during formal preference assessments (see Table 2).
Marcy was a bachelor’s student in special education and was one of the primary
therapists for Tommy. Marcy had a self-rating of 5 on the rapport Likert scale, and was chosen
100% of the time by Tommy during formal preference assessments (see Table 3). Marcy
participated during pre-intervention sessions with Tommy, but was unable to participate in post-
intervention sessions due to an employment change that occurred half way throughout the study.
Setting and Materials
During the rapport evaluation (i.e., pre-intervention and post-intervention conditions),
sessions were conducted in various locations at the university-based autism clinic that were
reported to be preferred for the child based on informal interviews with the child’s behavioral
19

therapists and during informal direct observations. The setting was intended to produce a natural
environment for social interactions. For Cole, all pre-intervention and post-intervention sessions
were conducted in his individual therapy room approximately 3.65 x 3.65 in dimension. The
room was equipped with two chairs, a table, and several age related and preferred toys that
remained in his room throughout the study. For Zane, pre-intervention and post-intervention
sessions were conducted in either the gym or the motor lab. The gym was approximately 18 m x
11 m in dimension, and was equipped with a trampoline, a large tunnel, two bicycles, two
scooters, and a Nintendo WiiTM
console game system, which was reflected on a large screen that
hung from the ceiling. The motor lab was approximately 7 m by 9 m in dimension, and was
equipped with a swing, a ball pit, two tables, four chairs, a trampoline, a bicycle, and several age
related toys (e.g., blocks, board games, etc.,). For Tommy, pre-intervention and post-
intervention sessions were conducted in the gym, motor lab (descriptions described above) or the
preschool classroom. The preschool classroom was approximately 8 m x 9 m in dimension, and
consisted of one large community table with four chairs, a carpeted area with various age related
activities and larger toys (e.g., blocks, foam letters, musical toys, cars, train set, etc.,), several
small tables with two chairs (intended for individual therapy), a sink, a coat rack, a radio, and
several shelves and cubbies containing various age appropriate smaller toys (e.g., bubbles, action
figures, small toy cars, etc.).
During the discrimination training evaluation, discrimination training (i.e., intervention)
was conducted in individual treatment rooms measuring approximately 3.65 m x 3.65 m in
dimension. Individual treatment rooms were equipped with a table, two chairs, a 12’’ x by 16’’
lunch tray containing five highly preferred (HP) edible and/or leisure items, and several low
preferred/neutral toys. Baseline and post-training sessions were designed to be analogous to the
20

natural environment and conducted in rooms that were anecdotally stated by therapists to be
preferred for a given child. For Cole, baseline and post-training sessions were conducted in his
individual treatment room. In addition to the stimuli/materials already discussed, baseline and
post-training sessions also contained response materials. This included response 1A materials (3
x 5 pink index card) or response 1B materials (8.5 x 11 paper with a blue “X” that extended from
each corner of the paper). For Zane, baseline and post-training sessions were conducted in either
the gym or motor lab. In addition to the stimuli/materials already discussed, baseline and post-
training sessions also contained response materials. This included response 2A materials (8.5 x
11 paper with a blue X that extended from each corner of the paper) or response 2B materials (an
empty tin can). For Tommy, baseline and post-training sessions were conducted in his classroom.
In addition to the stimuli/materials already discussed, baseline and post-training sessions also
contained response materials. This included response 3A materials (a yellow block in the shape
of a star and a 8.5 x 11 piece of paper with a vertical line in the center of the paper) or response
2B materials (3 x 5 pink index card).
Experimental Design
During the rapport evaluation, a multi-element design was used to evaluate and compare
the effects of different rapport levels (e.g., high rapport or lack of rapport) on behavioral
correlates of rapport. In addition, to evaluate whether discrimination training had an effect
rapport levels, pre-intervention and post-intervention conditions were conducted in a multiple
baseline design across dyads.
During discrimination training evaluations, baseline and post-training conditions were
conducted to assess whether social interaction (i.e., neutral stimulus) acquired reinforcing
properties as a result of discrimination training. A multiple baseline across dyads was used in
21

order to evaluate the effects of discrimination training on child-emitted independent responses.
In addition, probe tests were conducted throughout baseline and post-training conditions with a
second response (e.g., response “B”). The start of the discrimination training procedure was
staggered across the dyads after 4, 6, and 9 baseline data points.
Response Measurement
During rapport evaluations, pre-intervention and post-intervention sessions were
conducted with at least 30 minutes in-between each session. All sessions were video-recorded
and further analyzed by the primary experimenter and a second observer using a 5-second partial
interval recording method.
The primary dependent variable was the average percentage of child-emitted rapport
behaviors. Child emitted rapport behaviors include child approaches toward therapist, child eye
contact towards therapist, child body orientation facing therapist, child initiated physical contact,
and child smiles. In order to generate an average percentage of total rapport behaviors, each
target rapport behavior (e.g., approaches, eye contact, etc.,) was calculated as a percentage of
intervals for each session by adding the number of intervals with an occurrence of a target
rapport behavior, dividing by 36 (i.e., intervals per session) and multiplying by 100 to yield a
percent. Averages were further calculated by dividing the sum of percentages by the total
number of behaviors (e.g., five).
In addition, mutually emitted rapport behaviors and therapist emitted rapport behaviors
were recorded as a secondary measure. Therapist rapport behaviors included therapist
approaches towards child, therapist eye contact towards child, therapist body orientation towards
the child; therapist initiated physical contact towards the child, and therapist smiles. Mutual
rapport behaviors included responses exhibited by both the therapist and child simultaneously.
22

Mutual behaviors included mutual eye contact, mutual orientation, mutual physical contact, close
proximity, and mutual smiles. Average percentages of total mutual and total therapist emitted
rapport behaviors were calculated using the same method described above for child-emitted
rapport behaviors.
The average of child-emitted rapport behaviors was used as the primary dependent
measure to analyze trend within the data. Therefore, the average of child-emitted rapport
behaviors guided the experimenters’ decision on when to change phases (i.e., introduce
intervention sessions). All behaviors were operationally defined, and video-footage could be
watched as many times as needed, and often times, in slow motion, in order to measure the
occurrence or non-occurrence of a target behavior.
During discrimination training evaluations, baseline and post-training conditions were
intended to demonstrate the reinforcing properties of social interaction before and after
intervention. During baseline and post-training sessions, the primary dependent variable was the
rate of independent responses emitted by the child participant within a three-minute session. The
rate of independent responses was calculated by dividing the frequency of independent responses
by the total session time (i.e. 3 min).
During discrimination training sessions (i.e., intervention), the primary dependent
variable was the percentage of correct (i.e., unprompted) responses that occurred in each session.
During SD
trials, a correct response was scored if the child reached for the tray of highly
preferred stimuli without a physical prompt. During S-delta trials, a correct response was scored
if the child did not reach for a highly preferred item. The percentage of correct responses was
calculated by dividing the number of correct responses by 24 and multiplying by 100 to yield a
percentage.
23

In addition, data was collected on prompted responses and blocked responses. Blocks
were defined as physically guiding the child’s hand away from the tray of highly preferred items
during the S-delta trials. The experimenter only blocked the child attempts to reach for the tray
of preferred items during the S-delta trials. The percentage of blocked-responses was calculated
by dividing the number of blocks in a session by 12 and multiplying by 100 to yield a
percentage.
Prompts were defined as the experimenter physically guiding the child’s hand to pick up
the top ranked preferred item on the tray. Prompts were administered during SD
trials if 15s had
elapsed with no correct response. The percentage of prompted responses was calculated by
dividing the number of prompts by 12 and multiplying by 100 to yield a percentage.
Interobserver Agreement and Treatment Integrity
During rapport evaluations, a second observer recorded data from video-footage for a
minimum of 33% of sessions for each condition (i.e., pre-intervention and post-intervention) per
dyad (i.e., high-rapport and lack-of-rapport dyads). Interobserver agreement) was calculated on
an interval-by-interval basis by dividing the number of agreements by the total number of
agreements and disagreements and multiplied by 100 to yield a percentage. An agreement was
scored if both independent observers recorded an occurrence or nonoccurrence of a rapport
behavior during each interval. IOA was scored for each condition per dyad, and also across each
target rapport behavior (i.e., approaches, eye contact, body orientation, etc.,). Tables 5, 6 and 7
provide the mean agreement (i.e., IOA results) across dyads and target behaviors for Cole, Zane,
and Tommy respectively.
In addition, two independent observers recorded treatment integrity data for at least 44%
of pre-intervention and post-intervention sessions across each dyad during the rapport evaluation.
24

A checklist of necessary steps was used to score and calculate treatment integrity of the
experimenter and behavioral therapists. The following three steps were necessary: (1) Behavioral
therapist and child remained in location for 3-minutes, (2) Video footage was clear, with an
emphasis on child behaviors if both members of the dyad could not be captured, (3) staff did not
deliver demands (scored in 30-second partial interval recording). Both independent observers
scored 100% during treatment integrity checks for all sessions.
During discrimination training evaluations, a second observer recorded data from video
footage for a minimum of 33% of baseline and post-training sessions for Cole, Zane, and
Tommy. IOA was calculated using event recording by dividing the smaller number of observed
occurrences by the larger number of observed occurrences and multiplied by 100 to create a
percent. An occurrence was scored if an independent response was observed. Mean agreement
was 100% for all three participants during pre-test and post-test sessions.
During discrimination training sessions, two independent observers recorded data from
video footage for a minimum of 33% of discrimination training sessions for each dyad. IOA was
assessed for 60% of Cole’s discrimination training sessions, 33.33% of Zane’s discrimination
training sessions, and 36.36% of Tommy’s discrimination training sessions. IOA was calculated
on a trial-by-trial basis by dividing the number of agreements by the total number of agreements
plus disagreements and multiplied by 100. An agreement was scored if both independent
observers scored a correct response, prompt, or block for each trial. Mean agreement was 100%
across all three dyads.
In addition, integrity measures were collected. A second observer calculated treatment
integrity data for at least 44% of baseline and post-training conditions across each dyad. The
checklists (see Appendix B) involved six of the following necessary steps: (1) Therapist
25

physically prompts the target response (2) Therapist delivers brief social interaction (i.e., neutral
stimulus) contingent upon the first prompted response, (3) therapist physically prompts the target
response again, (4) therapist delivers brief social interaction contingent upon the second
prompted response, (5) therapist delivers brief social interaction contingent upon independent
(i.e., unprompted) target responses (if applicable), and (6) During omission of independent target
responses, the therapist provided no social interaction. Percentage of treatment integrity was
calculated by dividing the number of correctly implemented steps by the total number of possible
steps and multiplied by 100. Treatment integrity was 100% for each participant.
During discrimination training sessions, two independent observers assessed treatment
integrity from video footage for at least 44% of sessions for each participant. Integrity measures
were collected separately for both SD
and S-delta trials. A checklist of necessary steps was used
to calculate treatment integrity across 24 intervals. Appendix C provides the checklists and
corresponding data sheets for discrimination training conditions.
The checklist for SD
trials involved the following necessary steps: (1) Therapist presents
tray of child reinforcers, (2) therapist begins social interaction immediately, (3) therapist delivers
social interaction for about 15-seconds, (4) experimenter physically prompts child to reach for
the top ranked reinforcer if applicable, (5) experimenter says, “my turn” if applicable (i.e., for
Cole’s leisure items), (6) reinforcers are placed back on the tray, (7) therapist removes tray off
table, (8) therapist turns body completely around, and (9) experimenter remains behind child at
all times in a neutral stance.
The checklist for S-delta trials involved the following necessary steps: (1) Therapist
presents tray of child reinforcers, (2) therapist turns body completely around so that their back is
facing the child, (3) experimenter blocks using most-to-least intrusive prompting methods when
26

necessary, (4) therapist turns around to remove tray off table after approximately 15-seconds had
elapsed, (5) therapist removes tray off table, and (6) experimenter remains behind child at all
times in a neutral stance.
Treatment integrity was calculated on a trial-by-trial basis by dividing the number of
agreements by the total number of agreements plus disagreements and multiplied by 100. An
agreement was scored if both independent observers scored a correctly implemented step. Both
independent observers scored 100% for both SD
and S-delta treatment integrity measures.
Social Validity
At the conclusion of the study, a social validity questionnaire (see Appendix D) was
given to the three lack-of-rapport therapists who participated in discrimination training. The
questionnaire was completed by the therapists to assess their acceptability and perceived
effectiveness of the intervention.
27

CHAPTER 5
PROCEDURE
Pre-Intervention (Rapport Evaluation)
At the beginning of the first pre-intervention session, the experimenter cited a script to
the behavioral therapists: “Interact as you typically would with a child. Do not place demands.
This should be a fun 3-minutes.” The high-rapport dyad and lack-of-rapport dyad were observed
separately, with at least 30 minutes in-between sessions. All sessions were video- recorded and
further analyzed in 5-second intervals to capture the occurrence or non-occurrence of fifteen
target rapport behaviors (e.g., five child-emitted behaviors, five mutual behaviors, and five
therapist-emitted behaviors).
All three lack-of-rapport dyads met the criteria for intervention (i.e., lower and
differentiated levels of rapport behaviors compared to high-rapport dyads) and therefore,
participated in discrimination training evaluations. The high-rapport dyads only participated in
the rapport evaluations and did not participate in discrimination training evaluations.
Baseline (Discrimination Training Evaluation)
Baseline conditions were conducted in the child’s natural environment (i.e., outside of
treatment rooms) in preferred locations. The purpose of baseline sessions was to assess the
reinforcing properties of a therapist’s social interaction prior to intervention. In other words, the
purpose was to confirm that social interaction was a neutral stimulus.
Immediately prior to the start of the session, the lack-of-rapport therapist physically
prompted the child-participant to engage in two target responses in order for the child to
experience the contingencies in place. The target response was touching a 3 x 5 index card for
Cole, touching an “X” on a piece of paper for Zane, and moving a block in the shape of a square
28

across a line for Tommy. In addition, a second response was probed randomly throughout
baseline and post-training conditions. The probe responses were touching an “X” on a piece of
paper for Cole, tapping the top of a can for Zane, and touching a 3 x 5 index card for Tommy.
Contingent upon the first two prompted responses, the lack-of-rapport therapist delivered
brief (e.g., 10-15 s) social interaction. Social interaction typically consisted of praise statements
and a variety of nonverbal behaviors (e.g., smiles, physical touch, eye contact, etc.,). However,
the therapist’s were only instructed to provide social interaction and the behaviors comprising
the interaction were not detailed. Materials required for the participant to engage in an
independent response were always within arms reach of the child-participant. The lack-of-
rapport therapist did not interact with the child during the session unless an independent (i.e.,
unprompted) response was made. When an independent response was made, the therapist
delivered social interaction for about 10-15 seconds.
Discrimination Training (Discrimination Training Evaluation)
All training sessions were conducted in individual therapy rooms with minimal
distractions. The lack-of-rapport therapist sat across from the child participant. The experimenter
(first author of current study) remained behind the child at all times in a neutral stance in order to
block during S-delta trials (if necessary) or physically prompt during SD
trials (if necessary).
Twelve SD
and twelve S-delta trials were presented in an alternating order (i.e., SD
trial, S-delta
trial, SD
trial, S-delta trial, etc…).
During SD
trials, the lack-of-rapport therapist placed a brown lunch tray containing five
of the child’s highly preferred edibles and/or leisure items on the table and immediately began
delivering social interaction (i.e., the discriminative stimulus). Social interaction was delivered
for the entire duration of the trial and typically consisted of the therapist emitting relevant
29

rapport behaviors (i.e., eye contact, body orientation facing child, close proximity, approaches,
smiles, proximity) although they were never directly instructed as to how to provide social
interaction.
When the child reached for any item on the tray (even just one item), the trial was scored
as a correct response and they were allowed access to the leisure item or consumption of the
edible for 15 seconds. If the child did not reach for an item on the tray within 15 seconds, the
experimenter physically prompted the child to reach for the highest ranked edible or leisure item
and the trial was scored as incorrect.
The SD
trial was terminated after 15 s of access to the leisure item (e.g., Cole), or after all
edibles on the tray were consumed (e.g., Zane and Tommy) which varied in time depending on
child-participant. If some, but not all edibles were consumed, the trial was terminated following
a therapist’s question of, “do you want any more?” and the child responding, “no.” Due to
intrinsic variables (e.g., consumption time of edibles differed for each child), the SD
trial time
was variable across participants.
During the S-delta trials, the lack-of-rapport therapist delivered the tray containing five
highly preferred (HP) stimuli on the table. The lack-of-rapport therapist immediately turned
his/her body completely around so that the therapists back was always facing the child-
participant (i.e., the therapist back was the S-delta). No social interaction was delivered during
this time. If the child attempted to reach for the tray, the experimenter physical blocked the
attempt to assure the child did not receive access to the HP stimuli during the S-delta trials. If a
block occurred, the trial was scored as an incorrect. If the child did not reach for the tray during
an S-delta trial, the trial was scored as a correct response. The S-delta trials were always
terminated after 15 s had elapsed.
30

Discrimination training sessions were conducted one-to-three times a day with at least
one-hour in between sessions. Discrimination training sessions lasted an average of 11 minutes
for Cole & Jan, 11 minutes for Zane & Raquel, and 15 minutes for Tommy & Katie.
Post-Training (Discrimination Training Evaluation)
Post-training began following mastery of discrimination training. Discrimination training
sessions continued to occur following mastery and prior to post-training sessions in order to
strengthen the reinforcing properties of the therapist’s social interaction. Post-training procedures
were identical to baseline procedures. The lack of rapport therapist provided social interaction
contingent upon the first two physically prompted responses. After the second prompted
response, the 3-minute session began and social interaction was only delivered contingent upon
the child making an independent response.
Post-Intervention (Rapport Evaluation)
Sessions were identical to pre-intervention during the rapport evaluation. That is, at the
beginning of the first post-intervention session, the researcher cited the following script to the
behavioral therapists: “Interact as you typically would with a child. Do not place demands. This
should be a fun 3-minutes.” Both high-rapport and lack-of-rapport dyads were observed
separately. Sessions were terminated after 3-minutes had elapsed. Each session was video-
recorded and further analyzed for child-emitted rapport behaviors, mutual rapport behaviors, and
therapist-emitted rapport behaviors.
31

CHAPTER 6
RESULTS
Discrimination Training Evaluation
Figure 1 depicts the rate of responses emitted by each child participant (Cole, Zane &
Tommy) during baseline and post-training sessions. When the rate of responding remained stable
or was on a decreasing trend, discrimination training was conducted (see Figures 2, 3, and 4).
Baseline and post-training sessions for Cole are shown in the top panel of Figure 1. The
rate of responding during baseline remained relatively low for both response sets. For response
1A, the average rate of responding was 0.22 (range, 0 to 0.33) responses per minute. One probe
session was conducted for response set 1B during baseline, and Cole’s rate of responding was
zero responses per minute. Following discrimination training, post-training sessions were
conducted. For response 1A, Cole’s rate of responding immediately increased to 5 responses per
minute. The rate for response 1A was on average 3.66 (range, 2.0 to 5.0) responses per minute
across post-training sessions. For response 1B, Cole’s rate of responding increased to 1.66
responses per minute.
Baseline and post-training sessions for Zane are depicted in the second panel of Figure 1.
Zane’s rate of responding for response 2A was variable during baseline, but remained relatively
low, occurring at an average rate of 1.33 responses per minute (range, 0.33 to 2.66) across six
baseline sessions. Responses occurring in response set 2B were initially high during the first
baseline session (session 4), occurring at a rate of 4.33 responses per minute. The rate of
responses for 2B decreased to zero during baseline. The average rate was 1.88 responses per
minute for response set 2B during baseline. During post-training sessions, Zane’s responding
32

increased to an average of 3.44 responses per minute for response 2A. For response set 2B, the
rate increased to an average of 3.83 responses per minute during post-training sessions
Tommy’s data during baseline and post-training are depicted in the third panel of Figure
1. For both responses 3A and 3B, rate of responding remained relatively low during baseline
sessions at an average of .17 responses per minute for response 3A, and an average of .5
responses per minute for response 3B. During post-training sessions, the rate immediately
increased for response 3A and 3B. The average rate for 3A was 1.22 responses per minute with a
range of 0 to 2.33 responses per minute. One data point during post-training sessions was at zero
(e.g., session six), but this was most likely due to a confounding variable in the environment (i.e.,
a novel competing reinforcer). Tommy’s rate of responding during response 3A immediately
increased during post-training sessions (rate of 2 responses per minute), and continued on an
increasing trend.
Data during discrimination training is displayed in Figure’s 2-4. Cole (see Figure 2) and
Zane (see Figure 3) reached mastery criterion (i.e., minimum of 90% correct responding across
two consecutive sessions) during session number three. It took Tommy seven sessions before
reaching mastery criterion (see Figure 4). Discrimination training continued for all participants
throughout post-training sessions.
Rapport Evaluation
The results of the rapport evaluation can be seen in figures 5-7. Figure 5 displays the
average percentage of total rapport behaviors emitted by the child. The target rapport behaviors
that are measured include: child approaches towards therapist, child eye contact with therapist,
child body orientation towards therapist, child initiated physical contact to therapist, and child
smiles.
33

For Cole (see top panel of Figure 5), the average percentage of intervals with child-
emitted rapport behaviors was 36.67% (range, 33.89% to 38.33%) during pre-intervention
sessions with the high-rapport dyad (i.e., Cole & Rick) and 12.41% (range, 3.33% to 18.89%)
during pre-intervention sessions with the lack-of-rapport dyad (i.e., Cole & Jan). During post-
intervention sessions, the percentage of intervals with child emitted rapport behaviors increased
to an average of 34.63% (range, 26.66% to 45%) during lack-of-rapport sessions. This is a
22.22% increase of child-emitted rapport behaviors following intervention.
For Zane (see middle panel of Figure 5), the average percentage of intervals with child-
emitted rapport behaviors was 45.42% (range, 43.33% to 51.11%) across four pre-intervention
sessions with the high-rapport dyad (i.e., Zane & Ry) and 25.23% (range, 24.07% to 28.4%)
across four pre-intervention sessions with the lack-of-rapport dyad (i.e., Zane & Raquel). During
post-intervention sessions, the percentage of intervals with child emitted rapport behaviors
increased to an average of 55.56% (range, 48.89% to 61.67%) during lack-of-rapport sessions.
This is a 30.42% increase of target rapport behaviors following intervention for the lack-of-
rapport dyad compared to pre-intervention measures.
For Tommy, the average percentage of intervals with child emitted rapport behaviors was
33.67%% (range, 28.33% to 39.44%) across five pre-intervention sessions with the high-rapport
dyad (i.e., Tommy & Maci), and 12.33% (range, 9.44% to 20%) across pre-intervention sessions
with the lack-of-rapport dyad (i.e., Tommy & Katie). During post-intervention sessions, the
percentage of intervals with child emitted rapport behaviors increased to an average of 44.07%
(range, 38.33% to 53.33%) during lack-of-rapport sessions. This is a 31.74% increase of target
behaviors following intervention for the lack-of-rapport dyad compared to pre-intervention
measures.
34

Figure 6 displays the average percentage of mutual rapport behaviors (e.g., mutual eye
contact, mutual body orientation, mutual physical contact, close proximity, and mutual smiles)
exhibited by each dyad during pre-intervention and post-intervention conditions. For Cole, Zane,
and Tommy, the average percentage of intervals containing mutual rapport behaviors was
significantly higher during pre-intervention sessions with the high-rapport dyad (M= 60%,
45.28%, and 45%) compared to pre-intervention sessions with the lack-of-rapport dyad (M=
25.55%, 24.17%, and 22.22%). Following intervention (i.e., discrimination training), the average
percent of intervals with mutual rapport behaviors increased for Cole & Jan, Zane & Raquel, and
Tommy & Katie (M= 50.19%, 61.85%, and 59.07%) respectively.
Figure 7 displays the average percentage of therapist-emitted rapport behaviors (e.g.,
mutual eye contact, mutual body orientation, mutual physical contact, close proximity, and
mutual smiles) exhibited by each dyad during pre-intervention and post-intervention conditions.
For Cole & Jan, Zane & Raquel, and Tommy & Katie, the average percentage of intervals
containing therapist-rapport behaviors was significantly higher during pre-intervention sessions
with the high-rapport dyad (= 51.85%, 45.42%, and 52.33%) compared to pre-intervention
sessions with the lack-of-rapport dyad (M= 40%, 32.22%, and 34.56%). Following
discrimination training (i.e., post-intervention), the average percent of intervals with therapist
emitted rapport behaviors increased Cole & Raquel, Zane & Raquel, and Tommy & Katie (M=
57.78%, 61%, and 67.41%) respectively.
In addition to calculating average percentages of the combined target rapport behaviors,
the percentage of intervals with each child-emitted target rapport behavior for Cole, Zane, and
Tommy were calculated. The average percentage of intervals that Cole, Zane, and Tommy
engaged in each target rapport behavior (i.e., approaches, eye contact, body orientation, physical
35

contact, and smiles) can be seen in Figures 8 to 22. In terms of which behavior occurred most
frequently (i.e., central tendency) throughout pre-intervention and post-intervention conditions,
the percentage of intervals with each target behavior across both high-rapport and lack-of-rapport
therapists were calculated. The behavior(s) emitted most frequently will further be discussed for
each child participant. .
For Cole (see Figures 8 to 12), body orientation towards a therapist occurred most
frequently (on an a average of 62.27% of intervals) followed by eye contact (37.04%)
approaches (25.46%), physical contact (23.15%) and smiles (22.68%). In terms of the highest
occurring rapport behavior, Cole’s body was oriented toward the high-rapport therapist an
average of 80.56% of intervals and oriented towards the lack-of-rapport therapist on an average
of 25% of intervals during pre-intervention. During post-intervention, Cole’s body was oriented
toward the high-rapport therapist an average of 94.44% of intervals and body orientation towards
the lack-of-rapport therapist occurred on an average of 98.15% of intervals. In terms of the
second highest occurring behavior, Cole displayed eye contact during 75% of intervals with the
high-rapport therapist, and 11.11% of intervals with the lack-of-rapport therapist during pre-
intervention. During post-intervention, eye contact with the high-rapport therapist dropped to an
average of 24.07% of intervals, and eye contact with the lack-of-rapport therapist increased to an
average of 18.52%
For Zane (see Figures 13 to 17), body orientation towards a therapist occurred most
frequently (on an a average of 73.41% of intervals) followed by smiles (52.78%), eye contact
(47.42%), approaches (26.39%) and physical contact (8.53%). In terms of the highest occurring
rapport behavior, Zane’s body was oriented toward the high-rapport therapist an average of
72.22% of intervals and oriented towards the lack-of-rapport therapist on an average of 42.36%
36

of intervals during pre-intervention. During post-intervention, Zane’s body was oriented toward
the high-rapport therapist an average of 95.37% of intervals and body orientation towards the
lack-of-rapport therapist occurred on an average of 94.44% of intervals. In terms of the second
highest occurring behavior, Zane smiled during 67.36% of intervals during interactions with the
high-rapport therapist, and 23.61% of intervals with the lack-of-rapport therapist during pre-
intervention. During post-intervention interactions, Zane’s smiles occurred on an average of
64.81% of intervals with the high-rapport therapist and smiled during 60.19% of intervals with
the lack-of-rapport therapist.
For Tommy (see Figures 18 to 22), body orientation towards a therapist occurred most
frequently (on an a average of 60.94% of intervals) followed by smiles (46.49%), eye contact
(34.55%), physical contact (32.33%) and approaches (24.28%). In terms of the highest occurring
rapport behavior, Tommy’s body was oriented toward the high-rapport therapist an average of
69.44% of intervals and oriented towards the lack-of-rapport therapist on an average of 32.77%
of intervals during pre-intervention. During post-intervention, Tommy’s body was oriented
toward the lack-of-rapport therapist on an average of 85.18% of intervals. Post-intervention
measures were not conducted with the high-rapport therapist. In terms of the second highest
occurring behavior, Tommy smiled during an average of 40.56% of intervals during interactions
with the high-rapport therapist, and 12.22% of intervals with the lack-of-rapport therapist during
pre-intervention. During post-intervention interactions, Tommy’s smiling increased during
interactions with the high-rapport therapist to 37.96% of intervals during post-intervention.
37

CHAPTER 7
GENERAL DISCUSSION
The current study evaluated if rapport could be defined using behavioral correlates and if
rapport could be improved following implementation of a discriminative-stimulus conditioning
procedure. Our results have several indications: (1) children with autism exhibit differential and
higher levels of rapport behaviors (e.g., eye contact, approaches, etc.,) in the presence of a high-
rapport therapist compared to lack-of-rapport therapists, (2) a discriminative stimulus procedure
was successful in establishing the lack-of-rapport therapists’ social interaction as a
discriminative stimulus, (3) social interaction, after being established as discriminative stimulus,
acquired reinforcing properties, and (4) after social interaction acquired reinforcing properties (as
a result of the discriminative stimulus procedure), behavioral correlates of rapport increased with
all three lack-of-rapport dyads to levels similar (and sometimes higher) to that of high-rapport
dyads.
The results of the discriminative stimulus evaluation suggest that social interaction was
successfully established as a conditioned reinforcer. If reinforcement is defined by its initial
effect on behavior, then the results demonstrated by Cole, Zane, and Tommy during the
discriminative training evaluation support the notion that social interaction was indeed
established as a conditioned reinforcer. However, it is possible that conditioned reinforcers have
other properties, besides a reinforcing one, and it is these other roles (e.g., discriminative
properties), which produce the effects attributed to reinforcement (Kelleher & Gollub, 1962).
It is also possible that the reinforcing effect may diminish over time without
systematically fading primary reinforcement out. The results suggest that social interaction
alone, without additional pairings with primary reinforcers, increased responding. However, the
38

current study did not conduct follow-up measures to evaluate if social interaction would maintain
responding without intermittent pairings with primary reinforcers over time. Future research
should continue to examine the optimal method to condition social reinforcers for children with
autism and methods in which to promote maintenance of the effect.
Given the conceptualization of rapport discussed by Tickle-Degnen and Rosenthal (1987,
1990), the current study specifically defined rapport as a construct comprised by approaches, eye
contact, body orientation, physical contact, proximity, and smiles. It was found that when
independent observers were given brief (3-minute) video footage of an interaction between a
behavioral therapist and a child with autism, both independent observers were able to accurately
assess rapport by measuring the nonverbal behaviors stated above. In other words, both
independent observers reliably measured higher and differentiated levels of rapport behaviors
during interactions with high-rapport dyads compared to interactions with lack-of-rapport dyads.
The results support the effectiveness of the method utilized by McLaughlin and Carr
(2005) that was adapted in the current study to form high-rapport and poor-rapport dyads. In
other words, the use of staff self-ratings and preference assessment ratings made by the child
with disabilities may be an effective method to describe relationship quality between an
instructor and a learner. In addition, the results suggest that the target behaviors selected for
measurement in the current study (e.g., eye contact, smiles, etc.,) were indeed behavioral
correlates of rapport.
This supports previous research that has found an association between perceived rapport
and nonverbal behavioral correlates (e.g., Bernieri, 1988; Bernieri, Gillis, Davis, Grahe, 1996)
and suggests rapport can be defined via directly observable measures. Future research should
39

continue to examine the behaviors associated with rapport, and specifically if there is an
association between verbal behaviors and rapport.
Given that rapport was found to be readily observable, it suggests that interventions
aimed to enhance rapport can now be measured using an objective method. The current study
evaluated if conditioning social interaction (delivered from the lack-of-rapport therapist) as a
reinforcer for a child with autism would increase the level of rapport between the dyad.
Our results suggest that following a discrimination training procedure, rapport behaviors
emitted by the lack-of-rapport dyad increased to levels similar and sometimes higher to that of
the rapport behaviors emitted by the rapport-dyad. Although child emitted rapport behaviors
were observed as the primary unit of measurement in the current study, it was found that mutual
rapport behaviors and therapist emitted rapport behaviors increased as well. This supports the
notion that rapport is reciprocal process involving both members of the dyad.
It is possible that as the child rapport behaviors increased (e.g., child eye contact), it
served to reinforce (or provide an attention seeking bid for attention) to the therapist’s behavior.
This trend was seen in all child participants and their respected therapists. For example,
following discrimination training, child eye contact increased for Cole, Zane, and Tommy and
therapist eye contact concurrently increased for Jan, Raquel, and Katie (lack-of-rapport
therapists).
In addition to the objective success of the current evaluation, social validity results (see
Appendix E) indicate that the lack-of-rapport therapists perceived the intervention as meaningful
and effective. Jan (lack-of-rapport therapist for Cole), Raquel (lack-of-rapport therapist for
Zane), and Katie (lack-of-rapport therapist for Tommy) rated the importance, significance, and
advantages of the intervention as “very important” and stated they could easily understand the
40

intervention and why it was conducted. In addition, Jan anecdotally stated that she now had a
“bond” with Cole that she did not have prior to the start of intervention.
In sum, the current evaluation provides (1) an objective method to measuring rapport, (2)
support for the discrimination training procedure in establishing neutral social stimuli as
conditioned reinforcers, and (3) support that establishing a therapist’s social interaction as a
conditioned reinforcer may increase levels of rapport. Despite the success of both the
measurement of rapport and enhancement of rapport, several limitations are worth noting.
The first limitation was that pre-intervention measures of rapport behaviors for Cole and
Zane were not randomized and were conducted in an alternating order. This may decrease the
internal validity of our results, as the author may not have controlled for order effects. However,
rapport behaviors exhibited by the lack-of-rapport dyad and the high-rapport dyad were not
likely effected by the sequential order given that observations of each dyad were separated by a
minimum of 30-minutes, and sometimes, due to lack of availability of therapist and child,
sessions were separated across days at a time.
The second limitation of the current study was related to the practical use of measuring
rapport behaviors. In the current study, fifteen behaviors (e.g., child smiles, mutual smiles,
therapist smiles, child eye contact, mutual eye contact, therapist eye contact, etc.,) were scored
every five seconds for each session. Although this method allowed us to capture the occurrence
and nonoccurrence of the target rapport behaviors, this measurement system may not be useful
for clinical use. Future research may want to evaluate if rapport behaviors can be captured using
a less time-consuming data collection method (e.g., momentary time sampling) or limiting the
behaviors that are measured may provide a method for clinicians to monitor staff-client
interactions and intervene accordingly. However, there is a trade-off as limiting which rapport
41

behaviors to measure, for example, only observing approaches and smiles, may produce
inconclusive results.
The results of child emitted rapport behaviors indicated that each child participant
expressed rapport differently and some measures were found to be more indicative of rapport
than others. For example, during Cole’s rapport pre-intervention (i.e., baseline) sessions, child-
emitted body orientation was significantly differentiated and higher during interactions with the
high-rapport therapist compared to his interactions with the lack-of-rapport therapist. This
suggests that body orientation was one of the more significant indicators of a positive social
interaction than physical contact, in which both the lack-of-rapport and rapport dyad produced
low and similar levels during pre-intervention measures. In addition, body orientation and child
smiles seemed to be significant indicators of rapport for Zane and Tommy, whereas physical
contact remained at low levels during interactions with both the high-rapport therapists and lack-
of-rapport therapists. In sum, body orientation, eye contact, and smiles seemed to be the
strongest indicator of rapport whereas physical contact seemed be the lowest indicator of rapport.
However, only three children were evaluated, all three in which had a diagnosis of autism
spectrum disorder (ASD). Therefore, although the current study may provide useful information
on the behavioral correlates of rapport displayed by children with autism, future research may
examine these behavioral correlates with populations aside from individuals with ASD.
The third limitation important to discuss is that the independent observers measuring
behavioral correlates of rapport were not blind observers. That is, the independent observers
were aware of whether the interaction being observed was a perceived high rapport or lack-of-
rapport dyad. Although high measures of interobserver agreement were found between two
independent observers, there still may be bias in the measurement system. Future research may
42

further validate these findings by utilizing independent observers who are blind to the
experimental conditions.
Table 1
Formation of Dyads (Cole)
Note. Dashes indicate no session’s run, criteria already met. Only the lack-of-rapport dyad
participated in the discrimination training evaluation. Both high-rapport and lack-of-rapport
dyads participated in the rapport evaluation
a
Comparison dyad (high-rapport). b
Intervention dyad (lack-of-rapport)
Table 2
Formation of Dyads (Zane)
a
Staff
Self-Rating
(Likert 0-5)
Child Rating on
Preference
Assessment
(# chosen/trials) Dyad Group
Ricka
5 3/3 High-rapport dyad
Stan 5 2/3 -----
Janb
1 0/3 Lack-of-rapport dyad
Tessa 2 1/3 -----
Staff
Self-Rating
(Likert 0-5)
Child Rating on
Preference
Assessment
Rya
Matt 4 2/3 -----
Raquelb
Katie 3 1/3 -----
43

Table 3
Formation of Dyads (Tommy)
a
Table 4
Results of Response Assessments
Note. Response A and response B met selection criterion (i.e., average rate of 1.0 responses per
minute or below across three consecutive sessions) and were selected for use during baseline and
Staff
Self-Rating
(Likert 0-5)
Child Rating on
Preference
Assessment
Marcya
Tessa 5 2/3 -----
Katieb
Raquel 3 0/3 -----
Child
Participant Response A
Average across
Three Sessions
(Responses/per
minute) Response B
Average across
Three Sessions
(Responses/per
minute)
Cole 1A: Tapping a
3x5 pink index
card
0 1B: Tapping a
blue “X” on
8.5x11 paper
0
Zane 2A: Tapping a
blue “X” on
8.5x11 paper
0.33 2B: Tapping
the top of an
empty can
0.11
Tommy 3A: moving a
star (i.e., block)
across a
centered line on
8.5x11 paper
0 3A: Tapping a
3x5 pink index
card
0.11
44

post-training sessions in the discrimination training evaluations. Cole’s responses are denoted as
response 1A and 1B. Zane’s responses are denoted as response 2A and 2B. Tommy’s responses
are denoted as response 3A and 3B. Responses that did not meet selection criterion were
discarded from further use in the study and are not depicted.
Table 5
Interobserver agreement results: Cole’s rapport evaluation
Note. Mean interobserver agreement (IOA) for Cole during rapport evaluation. Table depicts
IOA across high-rapport dyad, lack-of-rapport dyad, and target behaviors during pre-intervention
and post-intervention.
High-rapport dyad Pre-Intervention Post-Intervention
Child rapport behaviors 93.33% 93.33%
Mutual rapport behaviors 90.28% 96.11%
Therapist rapport
behaviors
93.61% 91.67%
Lack-of-rapport
dyad
Mutual rapport behaviors 88.83% 95%
Therapist rapport
behaviors
92.78% 90%
Target rapport
behaviors
Approaches 93.31% 84.03%
Eye contact 90.12% 94.44%
Body orientation 93.52% 99.08%
Physical contact 93.83% 93.06%
Proximity 90.74% 97.22%
Smiles 91.98% 91.20%
45

Table 6
Interobserver agreement results: Zane’s rapport evaluation
Note. Mean interobserver agreement (IOA) for Zane during rapport evaluation. Table depicts
and post-intervention.
Therapist rapport
behaviors
90% 90.55%
Lack-of-rapport
dyad
Mutual rapport behaviors 88.83% 90%
Therapist rapport
behaviors
92.78% 90.28%
Target rapport
behaviors
Body orientation 89.35% 100%
Proximity 88.89% 100%
Smiles 90.51% 93.06%
46

Table 7
Interobserver agreement results: Tommy’s rapport evaluation
Note. Mean interobserver agreement (IOA) for Tommy during rapport evaluation. Table depicts
and post-intervention. Dashes indicate sessions that were not conducted.
Child rapport behaviors 88.33% ----
Mutual rapport behaviors 92.22% ----
Therapist rapport
behaviors
91.94% ----
Lack-of-rapport
dyad
Child rapport behaviors 92.5% 90%
Therapist rapport
behaviors
90.83% 90.56%
Target rapport
behaviors
Body orientation 88.43% 93.52%
Proximity 90.97% 94.44%
Smiles 89.35% 90.28%
47

Baseline Post-Intervention
Figure 1. Rate of responding for each child participant. The double phase change lines indicate
discrimination training occurred (see Figures 2-4). Black circles denote response A. White
circles denote response B probes.
0
1
2
3
4
5
0 1 2 3 4 5 6 7 8 9 10
Rateofresponding
Independent Responses (1A)
Independent Responses (1B)
Cole
0
1
2
3
4
5
0 1 2 3 4 5 6 7 8 9 10
Rateofresponding
Zane
0
1
2
3
4
5
0 1 2 3 4 5 6 7 8 9 10
Rateofresponding
Session (3-minutes)
Tommy
48

Figure 2. Discrimination training for Cole. Mastery criterion denoted by red squares.
Figure 3. Discrimination training for Zane. Mastery criterion denoted by red square.
Figure 4. Discrimination training for Tommy. Mastery criterion denoted by red square.
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11
PercentageofResponses
Session (24-trial block)
Correct
Blocked
Prompted
Cole
Mastery
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11
Correct
Blocked
Prompted
Zane
Mastery
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11
Correct
Blocked
Prompted
Tommy
Mastery
49

Child Emitted Rapport Behaviors
Pre-intervention Post-intervention
Figure 5. Average percentage of intervals with child emitted rapport behaviors. Target child
exhibited rapport behaviors include: Approaches, eye contact, orientation, physical contact, and
smiles. Black diamonds depict rapport-dyad interactions. White diamonds depict lack-of-rapport
dyad interactions. Double phase change lines indicate that discrimination training occurred.
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
high-rapport dyad
lack-of-rapport dyad
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
AveragePercentageofIntervals
Session (3-minutes)
Cole
Zane
Tommy
50

Mutual Rapport Behaviors
Figure 6. Average percentage of intervals with rapport behaviors emitted by both the child and
therapist at the same time. Target mutually exhibited rapport behaviors include: eye contact,
orientation, physical contact, proximity, and smiles. Black diamonds depict rapport-dyad
interactions. White diamonds depict lack-of-rapport dyad interactions. Double phase change
lines indicate that discrimination training occurred
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
High-rapport dyad
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Session (3-minutes)
Cole
Zane
Tommy
51

Therapist Emitted Rapport Behaviors
Figure 7. Average percentage of intervals with therapist emitted rapport behaviors. Target
therapist exhibited rapport behaviors include: Approaches, eye contact, orientation, physical
contact, and smiles. Black diamonds depict rapport-dyad interactions. White diamonds depict
lack-of-rapport dyad interactions. Double phase change lines indicate that discrimination training
occurred
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
high-rapport dyad
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Session (3-minutes)
Tommy
Zane
Cole
52

Figure 8. Percentage of intervals with child
emitted approaches towards therapist
body orientation facing therapist
Sessions (3-minute)
smiles
emitted eye contact with therapist
Sessions (3-minute)
initiated physical contact with therapist
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12
Child Approaches
Pre-intervention Post- intervention
high-rapport dyad
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12
Child Body Orientation
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12
Child Smiles
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12
Child Eye Contact
Pre-intervention Post-Intervention
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12
Child Physical Contact
PercentageofIntervals
Cole
Child Emitted Target Rapport
Behaviors
lack-of-rapport
dyad
high-rapport dyad
lack-of-rapport
dyad
53

Sessions (3-minute)
smiles
Sessions (3-minute)
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Child Approaches
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Child Smiles
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Child Eye Contact
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Zane
Behaviors
lack-of-rapport
dyad
lack-of-rapport
dyad
high-rapport
dyad
high-rapport
dyad
54

Sessions (3-minute)
Figure 22. Percentage of intervals with
child smiles
Sessions (3-minute)
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Child Approaches
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Child Smiles
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Child Eye Contact
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Tommy
Behaviors
lack-of-rapport
dyad
lack-of-
rapport dyad
high-rapport
dyad high-rapport
dyad
55

APPENDIX A
TREATMENT INTEGRITY CHECKLIST FOR RESPONSE ASSESSMENT
56

Treatment Integrity Checklist
Reinforcer Assessment
Child: _______________Staff: ___________ Session: ________
Date: _________________ Target response: __________
Trial presentation Child
Response
1 First target response is
physically prompted by
experimenter
+ --
2 Second target response is
physically prompted
correctly by experimenter
+ --
3 No social interaction,
feedback, or instructions is
provided throughout the
session (omission of
interaction)
+ --
4 The opportunity to emit the
response is always present
(response materials within
close proximity to child)
+ --
Total # Correct (+)
Total # incorrect (-)
Total:
Correct/ (correct +
incorrect)
57

APPENDIX B
TREATMENT INTEGRITY CHECKLIST FOR BASELINE & POST-TRAINING SESSIONS
58

Treatment Integrity Checklist
Baseline or Post-training condition (circle)
Child: _______________Staff: ___________ Session: ________
Date: _________________ Target response: __________
Trial presentation Child
Response
1 Therapist physically prompts
the target response + --
2 Therapist delivers brief social
interaction/praise + --
3 Therapist physically prompts
the target response + --
4 Therapist delivers brief social
interaction/praise + --
5 If independent responses
occur, therapists delivered
brief social interaction
+ --
N/A
6 During omission of target
response, no social interaction
(NS) is delivered
+ --
Total # Correct(+)
Total # incorrect (-)
Total:
Correct/ (correct +
incorrect)
59

APPENDIX C
TREATMENT INTEGRITY CHECKLIST FOR DISCRIMINATION TRAINING
60

Treatment Integrity Checklist for SD
Trials
Condition: discrimination training procedure
Child: _________________ Staff: ____________ Prompter: ____________
Date: ________________ Session: ________
Trial presentation 1 3 5 7 9 11 13 15 17 19 21 23
1 Behavioral therapist presents
tray of child’s reinforcers
2 Therapist begins social
interaction immediately
(within 0-2 seconds)
3 Therapist delivers social
interaction for 15 seconds (+
or – 5 seconds)
4 Experimenter physically
prompts child to reach for #1
ranked SR+ when necessary
5 Experimenter says, “my turn”
6 SR+ are placed back on tray
(by therapist or experimenter)
7 Therapist removes tray off
table
8 Therapist turns body
completely around
9 Experimenter remains behind
child at all times in neutral
stance
61

Treatment Integrity Checklist for S-Delta Trials
Condition: discrimination training procedure
Child: _________________ Staff: ____________ Blocker: ____________
Date: ________________ Session: ________
Trial presentation 2 4 6 8 10 12 14 16 18 20 22 24
1 Behavioral therapist presents
tray of child’s reinforcers
2 Therapist turns body
completely around so she/he is
not facing child
3 Experimenter blocks using
most-to-least intrusive
prompting methods when
necessary
4 Therapist turns around to
remove tray after 15 seconds
(+ or – 5 seconds)
6 Therapist removes tray off of
table
7 Therapist turns body around so
he/she is not facing child
8 Experimenter remains behind
child in a neutral stance at all
times
62

APPENDIX D
SOCIAL VALIDITY QUESTIONNAIRE
63

Social Validity Questionnaire Directions: Circle the best answer
1. Do you feel that the level of rapport between behavioral therapists and children
with autism is a socially important variable?
(Circle one: not important, a little important, quite important, very important)
2. Prior to the current study, how effective has training been in teaching you how to
build rapport with a child?
(Circle one: ineffective, a little effective, quite effective, very effective)
3. How would you rate the effectiveness of the intervention for this case?
(Circle one: ineffective, a little effective, quite effective, very effective)
4. How would you rate the importance, significance, and advantages, of the
intervention?
(Circle one: not important, a little important, quite important, very important)
5. How much time was taken per week for the intervention?
(Circle one: hardly any time, a fair bit of time, a lot of time, too much time)
6. Was the time taken to implement worthwhile? (i.e., did the ends justify the means?)
(Circle one: not worth the time, a little worth the time, worth the time, well worth the
time)
7. How intrusive was the intervention?
(Circle one: not intrusive, a little intrusive, somewhat intrusive, very intrusive)
8. Did you understand the intervention and why it was conducted?
(Circle one: not at all, understood some, understood most, understood easily)
9. Do you think you could implement the intervention on your own?
  (Circle one: not at all, with a lot of assistance, with a little assistance, easily)
10. How willing are you to do similar interventions in the future?
(Circle one: not at all, a little willing, quite willing, very willing)
11. How satisfied are you with the results of the intervention?
(Circle one: not satisfied, a little satisfied, satisfied, very satisfied)
64

APPENDIX E
SOCIAL VALIDITY RESULTS
65

Respondent: Jan
Question Rating Comments
1. Do you feel that the level of
rapport between behavioral
therapists and children with
autism is a socially important
variable?
Very Important
2. Prior to the current study, how
effective has training been in
teaching you how to build
rapport with a child?
A little effective Only been instructed to pair
myself with reinforcers
3. How would you rate the
effectiveness of the intervention
for this case?
Very effective After the intervention, the
child is interacting with me
more and I feel we now
have amazing “bond.”
importance, significance, and
advantages, of the intervention?
Quite important
5. How much time was taken per
week for the intervention?
A fair bit of time
6. Was the time taken to
implement worthwhile? (i.e.,
did the ends justify the means?)
Well worth the time
7. How intrusive was the
intervention?  
Not intrusive
8. Did you understand the
intervention and why it was
conducted?
Understood easily
9. Do you think you could
implement the intervention on
your own?
With a little
assistance
10. How willing are you to do
similar interventions in the
future?
Very willing
11. How satisfied are you with the
results of the intervention?
Very satisfied After intervention, my
relationship seems to have
changed with him.
66

Respondent: Raquel
variable?
Very important
A little effective
effectiveness of the
intervention for this case?
Quite effective
advantages, of the
intervention?
Very important
A fair bit of time
did the ends justify the
means?)
Worth the time
intervention?  
Not intrusive
conducted?
Understood easily
your own?
Easily
future?
A little willing
Very satisfied
67

d
Respondent: Katie
variable?
Very Important
A little effective
effectiveness of the
intervention for this case?
Very effective
advantages, of the
intervention?
Quite important
A fair bit of time
did the ends justify the
means?)
Well worth the time
intervention?  
Not intrusive
conducted?
Understood easily
your own?
With a little
assistance
future?
Very willing
Very satisfied
68

REFERENCES
Aronson, E. (1984). The social animal. New York: W.H. Freeman.
Babad, E., Bernieri, F., & Rosenthal, R. (1987). Nonverbal and verbal behavior of preschool,
remedial, and elementary school teachers. American Educational Research Journal, 24,
405-415.
Babad, E., Bernieri, F., & Rosenthal, R. (1989). Nonverbal communication and leakage in the
behavior of biased and unbiased teachers. Journal of Personality and Social Psychology,
56, 89-94.
Bernieri, F. (1988). Coordinated movement and rapport in teacher- student interactions. Journal
of Nonverbal Behavior, 12, 120- 138.
Bernieri, F. J., Gillis, J. S., Davis, J. M., & Grahe, J. E. (1996). Dyad rapport and the accuracy of
its judgment across situations: A lens model approach. Journal of Personality and Social
Psychology, 71, 110–129.
Bernieri, F., Resnick, J. S., & Rosenthal. R. (1988). Synchrony, pseudo- synchrony, and
dissynchrony: Measuring the entrainment process in mother-infant dyads. Journal of
Personality and Social Psychology, 54, 243-253.
Carr, E. G., McConnachie, G., Levin, L., & Kemp, D. C. (1993) Communication-based treatment
of severe behavior problems. (1993). In Behavior analysis and treatment (pp. 231-267).
Boston, MA: Springer US. DOI: 10.1007/978-1-4757-9374-1_10
Dinsmoor, J. A. (1995a). Stimulus control: Part I. The Behavior Analyst, 18, 51–68. Retrieved
from http://www.ncbi.nlm.nih.gov/pmc/journals/557/
Dinsmoor, J. A. (1995b). Stimulus control: Part II. The Behavior Analyst, 18, 253–269.
Retrieved from http://www.ncbi.nlm.nih.gov/pmc/journals/557/
69

Dunlap, G., Eno-Hieneman, M., Clark, S., & Childs, A. (1995). A preliminary exploration of
rapport development. Unpublished manuscript. University of South Florida, Tampa.
Dunlap, G., & Koegel, R.I., (1980). Motivating autistic children through stimulus variation.
Journal of Applied Behavior Analysis, 13, 619- 627. DOI: 10.1901/jaba.1980.13-619
Egan, G. (1975). The skilled helper: A model for systematic helping and interpersonal relating.
Monterey, Calif: Brooks/Cole Pub. Co. DOI: 10.1037/0013611
Fantino, E. (1977). Conditioned reinforcement. In W.K. Honig & J.E.R. Staddon (Eds.),
Handbook of operant behavior (pp. 313-339). Englewood Cliffs, NJ: Prentice Hall.
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 of Applied Behavior Analysis, 25, 491-498. DOI:
10.1901/jaba.1992.25-491
Gillis, J. S., Bernieri, F.J., & Wooten, E. (1995). The effects of stimulus medium and feedback
on the judgment of rapport. Organizational Behavior and Human Decision Processes, 63,
33-45.
Gollub, L.R. (1970). Information on conditioned reinforcement: A review of conditioned
reinforcement, edited by Derek P. Hendry. Journal of the Experimental Analysis of
Behavior, 14, 361-372.
Holth, P., Vandbakk, M., Finstad, J., Grønnerud, E.M., & Mari, J. (2009). An operant analysis of
joint attention and the establishment of conditioned social reinforcers. European Journal
of Behavior Analysis, 10, 143-158. Retrieved from http://www.ejoba.org/
Kelleher, R.T., & Gollub, L.R. (1962). A review of positive conditioned reinforcement. Journal
of the Experimental Analysis of Behavior, 5, 543-597.
70

Koegel, R. I., Dyer, K., & Bell, I. K. (1987). The influence of child preferred activities on
autistic children’s social behavior. Journal of Applied Behavior Analysis, 20, 243- 252.
DOI: 10.1901/jaba.1987.20-243
Leaf, R. & McEachin, J. (1999). A work in progress behavior: Management strategies and a
curriculum for intensive behavioral treatment of autism. New York, NY: DRL Books.
Lovaas, O. I., Freitag, G., Kinder, M. I., Rubenstein, B. D., Schaeffer, B., & Simmons, J. Q.
(1966). Establishment of social reinforcers in two schizophrenic children on the basis of
food. Journal of Experimental Child Psychology, 4, 109-125.
Maurice, C., Green, G., & Luce, S. C. (1996). Behavioral intervention for young children with
autism: A manual for parents and professionals. Austin, TX: Pro-ed.
McLaughlin, D. M., & Carr, E. G. (2005). Quality of rapport as a setting event for problem
behavior. Journal of Positive Behavior Interventions, 7, (2), 68-91. DOI:
10.1177/10983007050070020401
McConnell, S. R., (2002). Interventions to facilitate social interaction for young children with
autism: Review of available research and recommendations for educational intervention
and future research. Journal of Autism and Developmental Disorders, 32, 351- 372.
DOI: 10.1023/A:1020537805154
O’Toole, G. (2012). Communication: Core interpersonal skills for health professionals (2nd
ed.). Sydney: Churchill Livingstone/Elsevier.
Roberts, S. D., & Bouchard, K. R. (1989). Establishing rapport in rehabilitative audiology.
Journal of the Academy of Rehabilitative Audiology, 22, 67–73. Retrieved from
http://www.audrehab.org/jara.htm
71

Strain, P. S., & Shores, R. S. (1977). Social reciprocity: A review of research and educational
implications. Exceptional Children, 43(8), 526-530. Retrieved from
http://journals.cec.sped.org/ec/
Taylor, B. A., & Fisher, J. (2010). Three important things to consider when starting intervention
for a child diagnosed with autism. Behavior Analysis in Practice, 3, 52-53
Tickle-Degnen, L., & Rosenthal, R. (1987). Group rapport and nonverbal behavior.
Group Processes and Intergroup Relations, 9, 113-136.
Tickle-Degnen, L., & Rosenthal, R. (1990). The nature of rapport and its nonverbal correlates.
Psychological Inquiry, 1(4), 285-91. DOI: 10.1207/s15327965pli0104_1
White, S.W., Keonig, K., Scahill, L. (2007). Social skills development in children with Autism
Spectrum Disorders: A review of the intervention research. Journal of Autism and
Developmental Disorders, 37, 1858-1868. DOI: I 10.1007/s10803-006-0320-x
Williams, B. A. (1994). Conditioned reinforcement: Experimental and theoretical issues. The
Behavior Analyst, 17, 261-285.
72

lapin-ci

More Related Content

What's hot (10)

Viewers also liked (20)

Similar to lapin-ci (17)

lapin-ci