Unholz-Bowden2020_Article_CaregiverTrainingViaTelehealth

Journal of Behavioral Education (2020) 29:246–281
https://doi.org/10.1007/s10864-020-09381-7
REVIEW PAPER
Caregiver Training Via Telehealth on Behavioral Procedures:

A Systematic Review
Emily Unholz‑Bowden1 · Jennifer J. McComas1 · Kristen L. McMaster1 ·

Shawn N. Girtler1 · Rebecca L. Kolb1 · Alefyah Shipchandler1
Published online: 2 April 2020

© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Telehealth has been utilized to provide behavioral services to families with children
with autism spectrum disorder (ASD) and other disabilities. This systematic review
provides an update on current research pertaining to the use of telehealth to provide
behavior analytic-based services and train caregivers in implementing behavioral
procedures. This review also describes information on reported training components
and caregivers’ procedural fidelity. Empirical studies were collected from five data-
bases. Overall, the studies provide evidence of the utility of telehealth as a service
delivery model for providing behavior analytic-based services and for training car-
egivers to implement behavioral assessments and procedures. The authors discuss
potential considerations for developing training packages and training caregivers via
telehealth. Future research should use experimental methods to determine effective
components for training individuals via telehealth to use behavioral procedures with
good fidelity as well as to detect other factors that may influence procedural fidelity.
Keywords Telehealth · Autism spectrum disorder · Developmental disability ·

Applied behavior analysis · Parent training
Introduction
Individuals with autism spectrum disorder (ASD) and other disabilities have ben-
efited from behavioral interventions based on the principles of applied behavior
analysis (ABA) to shape appropriate behavior or teach new skills (Ferguson et al.
2019; Wong et al. 2015). Although ABA-based services are in demand among fami-
lies of children with disabilities, geography often poses a substantial barrier to gain-
ing access to such services. For example, in the United States, families who live in
* Emily Unholz‑Bowden
unhol003@umn.edu
1
Department of Educational Psychology, University of Minnesota, 56 E River Rd, Minneapolis,
MN 55455, USA
1Vol:.(1234567890)
3
Journal of Behavioral Education (2020) 29:246–281 247
urban areas or near universities tend to have easier access to ABA-based services,
whereas there is often less access to these services for families who live in rural
areas (Bulgren 2002; Ferguson et al. 2019; Murphy and Ruble 2012). In addition,
many countries outside of the United States do not have any access to ABA-based
services (Ferguson et al. 2019; Salomone et al. 2014).
One emerging model for delivering ABA-based services to families who live out-
side urban centers is telehealth, which is defined as the use of electronic technology
to provide individuals with clinical services (Puskin et al. 2006). Although the use of
telehealth to deliver ABA-based services is a relatively new approach, there is evi-
dence that parents find telehealth to be an acceptable service delivery mechanism. In
a study by Lindgren et al. (2016), parents who were provided ABA-based services
via telehealth reported similarly high ratings of treatment acceptability compared to
parents who received services at home with a behavior analyst on site and to parents
who received services in a regional clinic via telehealth. In addition, two recent liter-
ature reviews by Neely et al. (2017) and Tomlinson et al. (2018) examined studies in
which caregivers of individuals with intellectual and developmental disorders were
coached to implement behavioral interventions via telehealth and found that, overall,
caregivers reported high ratings of acceptability of being coached by behavior thera-
pists via telehealth. Considering the issue of limited access to ABA-based services
in many areas and evidence indicating telehealth is an acceptable option for service
delivery, further in-depth analysis of the use of telehealth is warranted.
A series of recent publications have demonstrated desirable ABA-based treat-
ment effects of behavioral coaching provided via telehealth for individuals with dis-
abilities (Barretto et al. 2006; Fisher et al. 2014; Lindgren et al. 2016; Wacker et al.
2013a). For example, Lindgren et al. (2016) compared outcomes of three groups of
children who were provided with ABA-based services: (1) at home with a behavior
analyst on site, (2) in a regional clinic with remote coaching via telehealth from a
behavior analyst, and (3) at home via telehealth with remote coaching from a behav-
ior analyst. Results indicated that all groups exhibited mean percentage reductions
in problem behavior of 90% or greater, and there were no significant differences in
problem behavior reduction among the three groups.
A number of recent empirical studies indicate that caregivers have successfully
implemented behavioral procedures they learned from behavioral therapists via tel-
ehealth (Neely et al. 2017) and some scholars have considered the “effectiveness”
of telehealth (Ferguson et al. 2019; Tomlinson et al. 2018). However, there is no
agreed-upon index of what aspect of telehealth should be targeted for examination
with regard to “effectiveness.” One literature review examined the extent to which
studies using telehealth met quality indicators for evidence-based practice and con-
cluded that much of the research on telehealth does not meet such quality indicators
(Ferguson et al. 2019). Using a quality assessment rubric (Reichow et al. 2008), the
authors determined that all group design studies included in the review were over-
all weak in quality and all single-case design studies were overall either “weak” or
“adequate” in quality. Similar results were also found in the Tomlinson et al. (2018)
review using the same rubric. However, two issues should be considered. First, it
is important to consider the areas in which the authors found lack of quality and
the extent to which those specific indicators affect the quality of the experimental
13
248 Journal of Behavioral Education (2020) 29:246–281
design and the demonstration of a functional relation or treatment effect. Second,

given that telehealth is not the independent variable in any of these studies, examin-
ing the quality indicators that support the effectiveness of telehealth is somewhat
misleading. Telehealth in itself is not an independent variable or an intervention.
It is a service-delivery mechanism, and the intervention component(s) and training
provided via telehealth are the independent variable(s).
With regard to analysis of the quality of studies involving telehealth, Ferguson
et al. (2019) assigned high ratings to very few single-case-design studies for present-
ing baseline data. They assigned high ratings to both group design and single-case-
design studies for their descriptions of the independent and dependent variables. The
indicators in both group and single-case-design studies that tended to result in lower
scores pertained to descriptions of (a) the participants and (b) comparison groups
(Ferguson et al. 2019). Although lack of such descriptions may preclude replica-
tion of a study, experimental control and treatment effects can still exist despite a
lack of detailed participant descriptions. In the Tomlinson et al. (2018) review, the
authors assigned low scores to single-case-design studies for presentation of stable
baseline data and other criteria related to presenting data that portray experimental
control. Additionally, they assigned low scores to single-case-design studies due to
lack of Kappa statistics, blind raters, or procedural fidelity measures. They assigned
low scores to group design studies for statistical analysis and sample size as well as
lack of blind raters, effect size measures, and procedural fidelity measures. Again,
although the presentation of stable baseline and intervention data in single-case-
design studies and the use of effect sizes in group design studies are important in
determining experimental rigor, experimental control, and replicability of a study,
the inclusion of Kappa statistics and blind raters in single-case-design studies as
well as large sample sizes in modified group design studies may not necessarily be
an indication of telehealth as being “ineffective.”
Despite the authors of these reviews concluding that the evidence supporting
telehealth is weak, the practice of training caregivers via telehealth should not be
discounted. It is important to note that telehealth is not an evidence-based practice
for addressing behavior change. Rather, it is a mechanism by which profession-
als can provide services and teach others to implement evidence-based practices.
Although meeting all quality indicators is an important priority within research to
provide strong evidence in support of particular assessments, interventions, and
training practices, studying those specific practices as they are delivered via tele-
health should be scrutinized when examining the utility of telehealth as a service
delivery mechanism. If the goal is to use telehealth to increase access to effective
services, an important question to be asked is: Under what conditions is training via
telehealth effective for teaching caregivers to implement ABA-based assessment and
interventions?
Therefore, one interest may be to examine training components used to teach
caregivers to implement behavioral procedures. Neely et al. (2017) reviewed the
various training components researchers have used most often while providing
training in behavioral procedures to caregivers via telehealth. The authors found
that, out of 19 studies involving the use of telehealth and training caregivers,
the most common training components used were performance feedback, vocal
13
instructions, and written instructions. Although feedback and instructions are the
most common training components, these components and others have been com-
bined to form a variety of training packages for use via telehealth. For exam-
ple, in studies by Dimian et al. (2018), Martens et al. (2019), and Wacker et al.
(2013a, b), researchers taught caregivers to implement functional communica-
tion training (FCT) with child participants. Although all caregiver participants
across all three studies were taught similar procedures, the researchers in each
study used vastly different treatment packages. In Wacker et al. (2013a, b), car-
egiver participants were provided with more structured and standardized train-
ing involving a variety of training components, such as a manual with extensive
information about ABA-based procedures used in the study, a formal presentation
on ABA-based principles and procedures, and vocal instructions during live ses-
sions. Dimian et al. (2018) also used a structured training package but with fewer
training components. Researchers used task analyses of the procedures imple-
mented as a guide for providing vocal instructions and performance feedback to
parents during live sessions. Martens et al. (2019) used minimal training com-
ponents, involving only vocal instructions to caregivers during live sessions and
prompting of specific steps when needed. Considering the spectrum of training
components and packages used in the literature, it may be difficult to identify the
necessary and sufficient methods for effectively training caregivers via telehealth.
Thus, when investigating the current literature, in addition to gathering informa-
tion regarding training components used, it might of interest to examine levels of
fidelity reported given the specified training components. Doing so may provide
preliminary information regarding promising approaches for training caregivers
via telehealth, which in turn may advance telehealth as a consistently effective
and generalizable service delivery mechanism.
Study Purpose
Previous literature reviews have provided a synthesis on literature related to car-

egiver training via telehealth with additional aims, such as providing information on
the acceptability of telehealth (Tomlinson et al. 2018), the methodological quality
of telehealth literature (Ferguson et al. 2019; Tomlinson et al. 2018), and the impact
of different training packages on the procedural fidelity of caregivers implementing
ABA-based procedures (Neely et al. 2017). The purpose of this literature review is
to (1) provide an update on current research pertaining to the use of telehealth to
provide ABA-based training to caregivers of individuals with ASD and other dis-
abilities and (2) document and examine training components used and procedural
fidelity levels reported across telehealth studies. Similar to Neely et al. (2017), we
aimed to learn more about potentially promising practices for implementing effec-
tive training to caregivers via telehealth. The current review contributes to the lit-
erature by synthesizing results of the literature that describes caregiver training via
telehealth and to spark future research pertaining to specific training components
and procedural fidelity.
13
Method
Search Procedures and Search Terms
In March 2019, the authors conducted a systematic literature search. We used

similar databases and search terms that Ferguson et al. (2019) and Neely et al.
(2017) used. The following databases were used: Academic Search Premier, Psy-
cINFO, Scopus, ERIC, and MEDLINE. Similar to Ferguson et al. (2019) and
Neely et al. (2017), results were restricted to peer-reviewed or academic arti-
cles and were not restricted based on year of publication. The terms used for the
search described individuals with ASD or a developmental disability and tele-
health. The terms were used in combination to identify articles that included both
descriptions. The search terms used which described individuals with disabili-
ties were Autis*, Asperger, “ASD”, “PDD-NOS”, and “Developmental Disab*.”
Terms used which described telehealth were telehealth, telepractice, telemedi-
cine, videoconferenc*, “distance education”, “distance train*”, teleconference,
telecare, “distance learn*”, and Elearn*.
The steps of the search are depicted in the PRISMA (Moher et al. 2009)
flowchart in Fig. 1. After articles were identified, the titles and abstracts were
screened. A full-text screening was then conducted on articles which were iden-
tified as relevant and met inclusion criteria described below through the initial
screening. To identify additional articles, we searched the reference lists of the
included articles. Reference lists from previous literature reviews were also
searched (Boisvert et al. 2010; Ferguson et al. 2019; Knutsen et al. 2016; Neely
et al. 2017; Sutherland et al. 2018; Tomlinson et al. 2018).
Inclusion Criteria
The following inclusion criteria were used to identify empirical studies to be exam-
ined in this synthesis: (1) the study involved caregiver or professional training which
was delivered exclusively through telehealth, (2) studies that involved child partici-
pants targeted children diagnosed with either ASD or a developmental disability, (3)
the training involved teaching caregivers (e.g., parents, teachers, students, behavioral
technicians) to use behavioral principles in treating problem behavior or teaching
skill acquisition with individuals with ASD, a developmental disability, or confeder-
ates, (4) the study included researcher-collected direct observation data depicting
assessment or intervention outcomes for child and/or caregiver participant behavior,
(5) the study involved teaching behavioral procedures to caregivers who had no prior
training in implementing those procedures, and (6) the study was written in English.
Across the five databases, 1159 total studies, excluding duplicates, were identi-
fied. After screening the titles and abstracts, 98 studies remained to be assessed
by their full text. From the 98 studies, 26 met eligibility to be included in the
review. Four additional studies were identified through search of reference lists.
In total, 30 studies were included in the synthesis.
13
Records identified through database

searching
(n = 1175)
Records after duplicates removed

(n =1159)
Records screened Records excluded

(n = 1159) (n =1061)
Full-text articles assessed Full-text articles excluded,

for eligibility with reasons
(n =98) (n =72)
Additional studies
Studies included in the
included from reference
synthesis
lists
(n =30)
(n =4)
Fig. 1 PRISMA flowchart depicting the literature search process
Data Extraction
For the synthesis, the 30 included studies selected were coded for the following
information: (1) child participant characteristics, (2) caregiver participant character-
istics, (3) research design, (4) behavioral procedures taught to caregivers to imple-
ment, (5) dependent variables involving direct observation of behavior exhibited by
caregiver participants, child participants, or both, (6) participant outcomes in rela-
tion to child participant behavior, caregiver participant skills in implementing ABA-
based procedures, or both, (7) whether sessions were conducted live or recorded by
caregivers, (8) training components used with caregiver participants, and (9) proce-
dural fidelity of caregiver participants. Some of the codes were similar to what was
included in Ferguson et al. (2019) and Neely et al. (2017).
13
Inter‑Rater Agreement
Search Agreement Procedures
A second reviewer used the same search terms described earlier in this paper to
obtain studies from the five databases listed. The second reviewer recorded the
number of studies obtained from each database. For each database, inter-rater
agreement (IRA) was obtained by dividing the smaller number with the larger
number and multiplying that result by 100. The average IRA across the five data-
bases was 94.1% (range from 92.5 to 96.3%).
Inclusion Agreement Procedures
Abstracts and titles from 30% (n = 348) of the studies gathered during the ini-
tial database search were reviewed by a second reviewer. For each study, the sec-
ond reviewer recorded whether the study met inclusion criteria. The final IRA
was calculated by dividing the number of agreements (331) by the sum of the
agreements plus disagreements (331 + 17) and multiplying that number by 100.
An IRA of 95.1% was obtained. Thirty percent (n = 29) of the studies chosen for
full-text review were also read by a second reviewer. For each study, the second
reviewer recorded whether the study met inclusion criteria. Questions about the
inclusion criteria were addressed by the first author. After clarifications were dis-
cussed, IRA was obtained by dividing the number of agreements (26) by the sum
of the agreements plus disagreements (26 + 3) and multiplying that number by
100. The resulting IRA was 89.7%.
Data Extraction Agreement Procedures
Lastly, 30% (n = 9) of the studies selected for the review were read by a second
reviewer to compare results coded by the primary reviewer. The second reviewer
coded for the same information described earlier in this paper. In total, the 9 arti-
cles read multiplied by nine codes recorded per article resulted in 81 opportuni-
ties for agreement. An iterative process was conducted in which code definitions
were reviewed in the context of several articles until an agreement on the coding
definitions and scheme was reached. The second reviewer made changes to her
codes on the nine articles and the primary reviewer independently recoded all
included studies according to the adjusted definitions. IRA was then calculated
by dividing the number of agreements (73) by the sum of the agreements plus
disagreements (73 + 8) and multiplying that number by 100. An IRA of 90.1%
was obtained.
13
Table 1 Qualitative synthesis of selected studies
Study Participants Research design Behavioral Dependent variables Outcomes Training components Procedural fidelity
procedures and session type
Alnemary Individual with a disability: Single case design: FA Teacher behavior: All teachers met criteria (a) Performance feed- Average during FA:
et al. one male child participated multiple baseline percentage of correct (90% or higher) on at back, (b) within- Teacher 1–90.5%
(2015)** during generalization across teachers with responses least two conditions of session instruction, Teacher 2–100%
Age: 12 an embedded multi- the FA (c) modeling, (d) Teacher 3–86.75%
Diagnosis: ASD element design One teacher met mastery pre-session instruc- Teacher 4–96.5%
Interventionists: four male criteria on all conditions tion, (e) address
teachers questions, (f) train-
One confederate included ing until meeting
Participants resided in Saudi criteria, (g) addi-
Arabia tional resources, (h)
booster training, (i)
role-play
Live sessions
Barkaia et al. Individuals with disabilities: Single case design: Mand and Child behavior: per- Increase in average levels (a) Performance Measured percentage of
(2017) three male children multiple baseline echoics centage of intervals of manding across two feedback, (b) pre- intervals with correct
Age: (M = 5.3) across participants training of mands and children and increase in session instruction, command sequencing
Diagnosis: ASD echoics average levels of echoics (c) written instruc- and positive conse-
Interventionists: three female Therapist behavior: across all children fol- tions, (d) prompt- quences
therapists percentage of lowing implementation of ing, (e) learning
Age: (M = 27) intervals of correct intervention checks
Participants resided in the command sequenc- Across therapists, increase Live sessions
country of Georgia ing and positive in average percentage of
consequences (pro- intervals of command
cedural fidelity) sequencing and positive
consequences following
coaching
253
13
Table 1 (continued)
254
13
Benson et al. Individuals with disabilities: Single case design: SDA; FA; FCT Child behavior: Results of assessments Within-session Parent 1:
(2018)** two male children ABAB; rates of SIB; rates of identified attention as instruction Average during FA-
Age: 5 and 8 Multielement design mands a reinforcer for one Live sessions 95%
Race: Caucasian for the FAs child and tangible as a Average during FCT-
Diagnosis: one with ASD; one reinforcer for the second 99%
with cerebral palsy child Parent 2:
Interventionists: Across both children, rates Average during SDA-
parents of child participants of SIB decreased and 81%
rates of mands increased Average during FA-
following the implemen- 77%
tation of FCT Average during FCT-
88%
Dimian et al. Individuals with disabilities: Single case design: SDA; FA; Child behavior: Child 1: (a) Performance Average during FCT:
(2018)** two male children multiple probe paired Child 1- Assessment indicated an feedback, (b) 92% and 96%
Age: 5.5 and 7 across commu- choice (a) percentage of escape function within-session
Race: Caucasian nicative contexts preference intervals of reach- Overall increase in AAC instruction, (c)
Diagnosis: multiple dis- with an embedded assessment; ing and pointing, symbol activation and written instructions
abilities reversal; FCT (b) percentage of decrease in tantrum and Live sessions
Interventionists: parents of multielement design intervals of crying reaching/pointing behav-
child 1; mother of child 2 for the SDAs and and screaming, (c) ior following implemen-
FAs occurrences of AAC tation of FCT
symbol activation Child 2:
Child 2-occurences Increase in AAC symbol
of AAC symbol activation following
activation implementation of FCT
Table 1 (continued)
Fischer et al. Individuals with disabilities: Single case design: Dyad 1: Child behavior: percent Across teachers, increase (a) Within-session Average during DRO:
(2017)** three children: one male; multiple baseline differential occurrences of in procedural fidelity fol- instruction, (b) 90%
one female; one unknown across teacher- reinforce- engagement; percent lowing consultations modeling, (c) Average during DRA:
Race: one Caucasian; one student dyads ment of occurrences of prob- Child 1: increase in engage- address questions, 97% across 2 par-
African American; one alternative lem behavior ment following imple- (d) MotivAiders ticipants
unknown behavior Teacher behavior: mentation of the GBG used for prompting, Average during GBG:
Diagnosis: one had ADHD (DRA) and procedural fidelity of Child 2: increase in engage- (e) review of data 76%
medication; one with ASD; the GBG procedures ment following imple- Live sessions
one unknown Dyad 2: DRA mentation of DRA
Interventionists: three female Dyad 3: Child 3: decrease in prob-
teachers differential lem behavior following
Race: Caucasian reinforce- implementation of DRO
ment of the
omission
of behavior
(DRO)
Fisher et al. Interventionists: one male and Group design: rand- Discrete- Technician behavior: All technicians improved in (a) Performance feed- Experimental group
(2014)** seven female behavioral omized clinical trial trial and procedural fidelity procedural fidelity scores back, (b) within- average:
technicians with waitlist control play-based scores on the Behav- on the post-test session instruction, BISWA-
Age: range from 21 to over 50 procedures ioral Implementation Large and statistically (c) modeling, (d) 97.5%
Study included confederates of Skills for Work significant mean differ- caregiver assess- BISPA-
instead of child participants Activities (BISWA) ence between treatment ment, (e) online 88.5%
and the Behavior and control group scores instruction, (f)
Implementation on BISPA and BISWA training until
of Skills for Play posttest meeting criteria, (g)
Activities (BISPA) learning checks, (h)
practice exercises,
(i) caregiver goal-
setting
Live sessions
255
13
Table 1 (continued)
256
13
Gibson et al. Individual with a disability: Single case design: FCT Child behavior: per- Decrease in elopement (a) Performance feed- Average during FCT:
(2010)** one male ABAB centage of intervals following implementation back, (b) within- 100% during baseline
Age: 4 of elopement of FCT session instruction, and 90% during
Diagnosis: ASD (c) modeling, (d) intervention
Interventionist: address questions,
one female teacher and one (e) written instruc-
teaching assistant tions, (f) training
until meeting
criteria, (g) collabo-
rative problem solv-
ing, (h) role-play
Live sessions
Hay-Hansson Individuals with disabilities: Group design: DTT Staff behavior: No significant differences (a) Performance feed- Measured percentage of
et al. six children randomized clinical procedural fidelity of in ETE scores at posttest back, (b) within- intervals with correct
(2013) Age: (M = 9.3) trial with random procedures according between groups who session instruction, implementation of
Diagnosis: ASD and/or assignment to an to scores on the received in-person or (c) modeling, (d) procedures
moderate developmental intervention Evaluation of Thera- remote training pre-session instruc-
disability peutic Effectiveness Overall, both groups tion, (e) caregiver
Interventionists: (ETE) significantly improved in assessment, (f)
three male and 13 female ETE scores from pre-test prompting, (g)
school staff to posttest practice exercises
Age: (M = 42.5) Live sessions
Table 1 (continued)
Heitzman- Interventionists: Group design: quasi- ABA-based Parent behavior: Across parents, increase (a) Performance Average posttest
Powell seven parents experimental with teaching procedural fidelity of in implementation of feedback, (b) performance across
et al. Age: (M = 37.3) pre- and posttest procedures procedures procedures from pre- to within-session participants and
(2014) and no control posttest instruction, (c) pre- skills: 71.8%
group session instruction,
(d) address ques-
tions, (e) caregiver
assessment, (f)
prompting, (g)
online instruction,
(h) written manual/
handouts, (i) train-
ing until meeting
criteria, (j) collabo-

rative problem solv-
ing, (k) learning
checks, (l) caregiver
reflections, (m)
practice exercises,
(n) performance
feedback on learn-
ing checks
Live sessions
257
13
Table 1 (continued)
258
13
Higgins et al. Individuals with disabilities: Single case design: Multiple stim- Technician behavior: Across technicians, increase (a) Performance feed- Technician 1 and 2:
(2017)** one female and two male multiple baseline ulus without procedural fidelity of in performance levels back, (b) modeling, 100% with confeder-
children across participants replacement procedures following implementation (c) pre-session ate and child during
Age: (M = 4.7) (MSWO) of training instruction, (d) post-training assess-
Diagnosis: ASD preference written instruc- ment
Interventionists: three female assessment tions, (e) caregiver Technician 3: 100%
behavioral technicians assessment, (f) with confederate and
Age: (M = 22.7) prompting, (g) 77.8% with child
online instruction, during post-training
(h) training until assessment; 92.3%
meeting criteria, (i) after tailored training
booster training, (j)
role-play, (k) video
self-modeling,
Live sessions
Table 1 (continued)
Ingersoll Individuals with disabilities: Group design: Naturalistic Child behavior: Children in both groups Self-directed group: Within 80-89% for both
et al. 19 male and eight female randomized clinical teaching rates of use of lan- exhibited significantly (a) modeling, (b) groups
(2016) children trial with random strategies guage targets higher use of language address questions,
Age: (M = 3.6) assignment to an Parent behavior: targets during posttest, (c) written instruc-
Race: 23 Caucasian; four intervention procedural fidelity of with the therapist-assisted tions, (d) caregiver
minorities procedures group scoring slightly assessment, (e)
Diagnosis: ASD or PDD-NOS higher online instruction,
Interventionists: Parents in both groups (f) written manual/
26 mothers and one father of exhibited significantly handouts, (g) learn-
the child participants higher procedural fidelity ing checks, (h) car-
during posttest, with the egiver reflections,
therapist-assisted group (i) practice outside
scoring significantly of sessions, (j) addi-
higher tional resources, (k)

homework assign-
ments
Additions to therapist-
assisted group:
(a) performance feed-
back, (b) within-
session instruction,
(c) pre-session
instruction
Live sessions
259
13
Table 1 (continued)
260
13
Knowles et al. Individuals with disabilities: Single case design: ABA-based Child behavior: per- Across children, decrease (a) Performance feed- Measured percentage of
(2017) four male children multiple baseline teaching centage of intervals in problem behavior back, (b) within- intervals with praise,
Age: 8-9 years across behaviors procedures of problem behavior following implementation session instruction, antecedent strategies,
Race: two African American; with four phases (praise, Teacher behavior: per- of training during phase (c) modeling, (d) and opportunities to
one Latino; one Pacific according to the prompting, centage of intervals one; continued to stay address questions, respond (OTR)
Islander behavioral proce- precorrec- of using procedures low across phases (e) online instruc-
Disability: two with EBD; two dures tion, and Average increase in teacher tion, (f) review
with OHI opportunities implementation of all of data, (g) video
Interventionist: to respond procedures follow- self-modeling
one female teacher (OTR)) ing implementation of Live sessions
Age: 23 training
Race: Caucasian
Machalicek Individuals with disabilities: Single case design: FA; Child behavior: per- Function of problem behav- (a) Performance feed- Average during FA:
et al. one male and two female AB and multiele- FCT; centage of intervals ior identified as escape back, (b) within- Parent 1 and 2–85%
(2016)** children ment for treatment DRA; of problem behavior for child 1, tangible for session instruction, Parent 3–95%
Age: (M = 11) effects and compari- differential child 2, and escape and (c) modeling, (d) Average across inter-
Race: Caucasian sons; multielement negative tangible for child 3 pre-session instruc- vention procedures:
Diagnosis: design for the FAs reinforce- Across children and tion, (e) address Parent 1–89%
Child 1 and 3- ASD ment of interventions, decrease questions, (f) Parent 2–74%
Child 2- multiple disabilities alternative in problem behavior fol- written instructions, Parent 3–93%
Interventionists: behavior lowing implementation of (g) prompting, (h)
one father and two mothers of (DNRA); intervention written manual/
the child participants antecedent- Child 1: problem behavior handouts, (i) train-
Race: same as children based strate- occurred the least during ing until meeting
gies DNRA criteria, (j) col-
Child 2 and 3: problem laborative problem
behavior was indis- solving, (k) practice
criminately low across exercises, (l) review
conditions of data, (m) graph-
ing sessions
Live sessions
Table 1 (continued)
Machalicek Individuals with disabilities: Single case design: FA Child behavior: Function of problem (a) Performance Not provided
et al. two female children multielement percentage of intervals behavior identified as feedback, (b)
(2009a) Age: 11 and 7 of problem behavior attention and escape for within-session
Race: one Hispanic; one both children instruction, (c)
Caucasian written instructions
Diagnosis: moderate intel- Live sessions
lectual disability
Interventionists:
two graduate students
Machalicek Individuals with disabilities: No experimental Paired choice Child behavior: Across children, prefer- (a) Performance 100% across graduate
et al. three male children design included preference number of choices ences were able to be feedback, (b) students
(2009b)** Age: (M = 4.7) assessment Graduate student identified within-session
Race: two Caucasian; one behavior: procedural All graduate students instruction, (c)
Asian American fidelity of procedures implemented the assess- written instructions,

Diagnosis: two with ASD; ment with 100% fidelity (d) practice outside
one with pervasive develop- of sessions
mental disorder and speech Live sessions
delay
Interventionists:
three graduate students
Machalicek Individuals with disabilities: Single case design: FA Teacher behavior: Across teachers, increase in (a) Performance feed- Average across teach-
et al. six children multiple baseline procedural fidelity of ability to conduct an FA back, (b) within- ers: 97%
(2010)** Age: (M = 6) across participants procedures following implementation session instruction,
Race: five Caucasian; one with an embedded of training (c) modeling, (d)
Asian American multielement design Some improvement during written instructions,
Diagnosis: five with ASD; for the FAs baseline observed (e) training until
one with language delays meeting criteria,
Interventionists: (f) additional
six female teachers resources, (g)
Age: (M = 27) booster training
Race: five Caucasian; one Live sessions
Chinse and Polish
261
13
Table 1 (continued)
262
13
Martens et al. Individuals with disabilities: Single case design: SDA; FA Child behavior: Function of problem behav- (a) Within-session Average during SDA:
(2019)** four male children multielement design percentage of intervals ior identified as escape instruction, Parent 1 and 2–100%
Age: (M = 6.5) for the FAs of problem behavior for child 1 and tangible (b) prompting Parent 3–80%
Race: Caucasian for child 4 Live sessions Parent 4–92%
Diagnosis: A possible tangible and Average during FA:
Child 1, 3 and 4- ASD escape function identified Parent 1–98%
Child 1-Lissencephaly for child 2 Parent 2–99%
Child 2- Rett syndrome A tangible function with a Parent 3 and 4–94%
Interventionists: possible escape function
parents of child participants identified for child 3
Less differentiated results
were obtained from the
SDA
Meadan et al. Individuals with disabilities: Single case design: Naturalistic Child behavior: Across children, increase in (a) Performance feed- Multiplied a quality
(2016) one female and two male multiple baseline teaching occurrences of social responses and initiations back, (b) within- rating by the rate of
children across strategies strategies communication ini- following implementation session instruction, strategy implemen-
Age: (M = 3) (modeling, tiations; percentage of strategies (c) modeling, tation
Race: two Caucasian; one mand-model, of opportunities with Across parents, increase in (d) pre-session
Middle Eastern time delay, social communica- rate and quality in imple- instruction, (e)
Diagnosis: ASD and envi- tion responses mentation of strategies address questions,
Interventionists: ronmental Parent behavior: following coaching (f) caregiver assess-
three mothers of child par- arrange- quality and rate in ment, (g) written
ticipants ment) implementing strate- manual/handouts,
Race: same as child partici- gies (h) training until
pants meeting criteria,
(i) collaborative
problem solving,
(j) caregiver reflec-
tions, (k) video
self-modeling
Live sessions
Table 1 (continued)
Schieltz et al. Individuals with disabilities: Single case design: FA; FCT Child behavior: Function of problem behav- (a) Within-session Average during FCT:
(2018)** one male and one female multiple baseline percentage of intervals ior identified as tangible instruction, (b) col- Parent 1–96%
child across participants; of problem behavior; for one child and tangible laborative problem Parent 2–45%
Age: 2 and 6 multielement design percentage of inter- and escape for the second solving
Diagnosis: ASD for the FAs vals of mands; per- child Live sessions
Interventionists: two mothers centage of intervals Across children, levels
of child participants of task completion of problem behavior
remained variable
Manding and task comple-
tion increased for one
child and was variable
for the second child fol-
lowing implementation
of FCT
Simacek et al. Individuals with disabilities: Single case design: SDA; FA; FCT Child behavior: Function of problem (a) Performance feed- Average during FCT:
(2017)** three female children adapted multiple Most-to-least occurrences of AAC behavior identified as back, (b) within- Parent 1–96%
Age: (M = 3.7) probe design across prompting requests occurrences escape for one child and session instruction, Parent 2–93%
Diagnosis: two with ASD; contexts; with time of idiosyncratic escape and tangible for (c) modeling, (d) Parent 3–94%
one with Rett syndrome ABAB design in the delay behavior another child written instructions
Interventionists: first tier; multiele- One child exhibited Live sessions
parents of child participants ment design during idiosyncratic behavior
the SDAs and FAs across contexts during
the SDA
Across children and tiers,
increased levels of AAC
requests and decreased
levels of idiosyncratic
behavior following
implementation of FCT
263
13
Table 1 (continued)
264
13
Suess et al. Individuals with disabilities: Single case design: FA; FCT Child behavior: Function of problem (a) Performance During FCT independ-
(2014)** three male children multielement design percentage of intervals behavior identified as feedback, (b) ent trials:
Age: (M = 2.9) with alternations of problem behavior; both escape and tangible within-session Parent 1–74%
Diagnosis: PDD-NOS between coached percentage of task for all children instruction, (c) pre- Parent 2–87.3%
Interventionists: trials (A) and inde- completion; percent- Across two children, prob- session instruction, Parent 3–78.1%
parents of child participants pendent trials (B); age of opportunities lem behavior was more (d) prompting, (e) During FCT coached
Age: (M = 37) multielement design with manding likely to occur during written manual/ trials:
for the FAs Parent behavior: coaching sessions handouts, (f) Parent 1–76.8%
procedural fidelity of No consistent differences practice outside of Parent 2–94.1%
procedures during in levels of manding and sessions, (g) review Parent 3–77.6%
FCT task completion across of data
session types Live sessions and
For two parents, procedural recorded sessions
fidelity was slightly submitted by the
higher during coaching participants
trials versus independ-
ent trials
Suess et al. Individuals with disabilities: Single case design: FA; FCT Child behavior: Function of problem (a) Performance feed- Not provided
(2016) three male and two female multiple baseline rates of problem behavior identified as back, (b) within-
children design across partic- behavior; percentage escape for three children session instruction,
Age: (M = 4.9) ipants; multielement of task comple- and escape and tangible (c) pre-session
Diagnosis: ASD design for the FAs tion; percentage of for one child instruction (d)
Interventionists: manding One child had inconclusive written instructions,
parents of child participants results (e) prompting, (f)
Across children, problem caregiver reflec-
behavior decreased and tions, (g) practice
levels of manding and outside of sessions,
task completion increased (h) review of data,
following implementation (i) homework
of FCT assignments
Live sessions
Table 1 (continued)
Sump et al. Interventionists: two male and Single case design: Antecedent Student behavior: Across students, increase (a) Performance Average across
(2018)** five female undergraduate multiple baseline instructional procedural fidelity of in skills following imple- feedback, (b) participants and
students across skills; strategies; procedures mentation of training modeling, (c) pre- skills: > 90%
Age: (M = 23) alternating treat- MSWO Training was equally effec- session instruction,
Study used adult confederates ments design used preference tive via telehealth as in (d) prompting,
instead of children to compare training assessment; person (e) training until
modalities conse- meeting criteria, (f)
quence- booster training, (g)
based role-play
strategies Live sessions
Vismara et al. Individuals with disabilities: Group design: P-ESDM (par- Child behavior: Overall, children in (a) Performance feed- Used a rating scale
(2018) 17 male and seven female randomized clinical ent model) rates of imitation; rates both groups exhibited back, (b) within- In the ESDM group,
children trial with random of spontaneous com- increased rates of imita- session instruction, five parents scored a
Age: (M = 2.4) assignment to an munication; rates tion, with higher levels in (c) modeling, (d) 4 or above during the
Race: four Hispanic; 20 non- intervention of joint attention the P-ESDM group pre-session instruc- post-test
Hispanic behavior No significant differences tion, (e) caregiver In the comparison
Diagnosis: ASD Parent behavior: in rates of spontaneous assessment, (f) group, two parents
Interventionists: procedural fidelity of communication and joint online instruction, scored a 4 or above
five fathers and 19 mothers of procedures attention behavior across (g) collaborative during the post-test
child participants groups problem solving,
Following coaching, proce- (h) learning checks,
dural fidelity increased in (i) caregiver reflec-
both groups, with statisti- tions, (j) additional
cally significantly higher resources, (k) car-
fidelity in the P-ESDM egiver goal-setting
group Live sessions
265
13
Table 1 (continued)
266
13
Vismara et al. Individuals with disabilities: Single case design: P-ESDM Child behavior: Across children, average (a) Performance feed- Used a rating scale
(2013) eight children multiple baseline (parent- rates of functional rate of vocalizations back, (b) modeling, Average rating across
Age: (M = 2.3) across participants model) vocalizations; rates increased while rate of (c) pre-session participants during
Race: six Caucasian; one of joint attention joint attention initiations instruction, (d) car- intervention: 3.68
Latino; one Hispanic initiations remained the same fol- egiver assessment,
Diagnosis: ASD Parent behavior: lowing implementation (e) online instruc-
Interventionists: procedural fidelity of of ESDM tion, (f) written
one father and seven mothers procedures; levels of Across parents, increase in manual/handouts,
of child participants parent engagement procedural fidelity fol- (g) collaborative
Race: same as child partici- lowing coaching problem solving,
pants Across parents, increase (h) learning checks,
in level of engagement (i) caregiver reflec-
following ESDM imple- tions, (j) practice
mentation outside of sessions,
(k) additional
resources, (l) prac-
tice exercises
Live sessions
Table 1 (continued)
Vismara et al. Individuals with disabilities: Single case design: ESDM Child behavior: Across children, statisti- (a) Performance feed- Used a rating scale
(2012) one female and eight male multiple baseline rates of spontaneous cally significant overall back, (b) within- Average rating across
children across participants vocalizations and increase in spontaneous session instruction, participants during
Age: (M = 2.3) imitation vocalizations and imita- (c) modeling, (d) follow-up: 4.29
Diagnosis: six with ASD; Parent behavior: tion following ESDM address questions,
three with PDD-NOS procedural fidelity of implementation (e) caregiver
Interventionists: procedures Across parents, statisti- assessment, (f)
nine parents of child partici- Levels of parent–child cally significant higher online instruction,
pants interaction procedural fidelity levels (g) collaborative
Race: eight Caucasian; one following implementation problem solving,
Hispanic of coaching (h) learning checks,
Across parent–child dyads, (i) caregiver reflec-
increase in level of inter- tions, (j) practice
actions following ESDM outside of sessions,

implementation (k) additional
resources, (l) prac-
tice exercises, (m)
booster training
Live sessions
Wacker et al. Individuals with disabilities: Single case design: FA Child behavior: Function of problem behav- (a) Performance feed- Average across parents:
(2013a)** 20 children multielement design percentage of intervals ior identified as escape back, (b) within- 96% without cor-
Age: (M = 4.4) of problem behavior and tangible for 13 session instruction, rections; 97% with
Diagnosis: 13 with PDD- children, escape for two (c) pre-session corrections
NOS; seven with ASD children, and tangible for instruction, (d)
Interventionists: one father three children address questions,
and 19 mothers of child For two children, function (e) prompting, (f)
participants was not identified written manual/
Age: (M = 34) handouts, (g)
review of data
Live sessions
267
13
Table 1 (continued)
268
13
Wacker et al. Individuals with disabilities: Singe case design: FCT Child behavior: per- Across children, decrease (a) Performance feed- Not provided
(2013b) one female and 16 male multiple baseline centage of intervals in problem behavior back, (b) within-
children across participants of problem behavior following implementation session instruction,
Age: (M = 4.3) of FCT (c) modeling,
Diagnosis: seven with ASD; (d) pre-session
10 with PDD-NOS instruction, (e)
Interventionists: 16 mothers address questions,
and two fathers of child (f) written instruc-
participants tions, (g) written
Age: (M = 33) manual/handouts,
(h) practice outside
of sessions
Live sessions
Table 1 (continued)
Wainer and Individuals with disabilities: Single case design: Reciprocal Child behavior: rates During self-directed condi- (a) Performance feed- Used a rating scale
Ingersoll five children multiple baseline imitation of spontaneous tion, child 1, 2, and 3 back, (b) modeling,
(2015) Age: (M = 3.5) across participants training imitation exhibited an increase in (c) address ques-
Race: one Caucasian; two (RIT) Parent behavior: imitation tions, (d) caregiver
Asian; one Hispanic; one (contingent procedural fidelity of During coaching condition, assessment, (e)
multi-racial imitation, procedures child 1, 2, and 4 exhibited online instruc-
Diagnosis: ASD linguistic an increase in imitation tion, (f) written
Interventionists: mapping, During self-directed condi- manual/handouts,
five parents of child partici- modeling, tion, parent 1, 3, 4, and (g) collaborative
pants prompting, 5 achieved moderate to problem solving,
Race: same as child partici- reinforce- high levels of fidelity (h) learning checks,
pants ment, During coaching condi- (i) caregiver reflec-
All participants resided in pacing, tion, parent 1, 2 and 3 tions, (j) additional
Canada and child exhibited increased levels resources (k)

spontaneous of fidelity practice exercises,
imitation) (l) homework
assignments, (m)
performance feed-
back on learning
checks, (n) sum-
mary of content
Live sessions
Recorded sessions
submitted by one
parent
269
13
Table 1 (continued)
270
13
Wilczynski Individual with a disability: No experimental Naturalistic Child behavior: Child initiation of (a) Performance Average across skills:
et al. one male child design: case study teaching percentage of com- compliance increased feedback, (b) 95%
(2017)** Age: 5 strategies pliance during teacher’s use of within-session
Diagnosis: ASD Teacher behavior: naturalistic teaching instruction,
Interventionist: one female procedural fidelity of strategies. Completion of (c) address questions,
teacher procedures compliance remained the (d) caregiver
same (100%) assessment, (e)
Teacher implemented four online instruction,
out of seven strategies (f) learning checks,
with higher fidelity dur- (g) video self-
ing the post-training than modeling
during pretraining Recorded sessions
Fidelity on two of the submitted by the
strategies stayed the participant followed
same (100%) and fidel- by coaching and
ity decreased for one feedback sessions
strategy with a behavior
analyst
Double asterisks next to author names indicate studies that reported procedural fidelity in the form of a percentage of steps completed accurately and of at least 90% for at least
one caregiver participant
Results
All information described below can be found in detail in Table 1.
Participant Characteristics
In total, across the 30 reviewed studies, 27 included children with disabilities as

participants with a total of 157 child participants across the studies. Out of the
27 studies, 22 (81.5%) reported the gender of the child participants. There were
94 males and 30 females. Twenty-six (96.3%) studies reported ages of the child
participants, which ranged from 1 to 16 years. Twenty-five (92.6%) studies pro-
vided either the exact age or the mean age of child participants, with the mean
age across those studies being 5.4 years. One study reported the age range of four
children as being within eight to nine years. Fourteen studies (51.9%) reported
race/ethnicity of the child participants. Across these studies, 52 children were
Caucasian, 20 non-Hispanic, seven Hispanic, four Asian American, three Afri-
can American, two Latino, one Pacific Islander, one Middle Eastern, and one
multi-racial. Four participants were described as being minorities. Of the 157
child participants, 87 (55.4%) were diagnosed with ASD. Twenty-seven partici-
pants (17.2%) were described as having ASD or a pervasive developmental dis-
order (PDD-NOS), six (3.8%) with either ASD and/or a moderate developmental
disability, and one with ASD and Lissencephaly. Twenty-three percent (n = 36)
were identified as having other disabilities, such as PPD-NOS, Rett syndrome,
and emotional behavioral disorder (EBD). One participant was described as being
prescribed with ADHD medication.
In total, at least 191 caregiver participants, who served as interventionists,
were included across the 30 reviewed studies. The term “at least” is used because
some studies did not include a specified number of caregivers who participated
in the studies. Of the 30 studies, 20 (66.7%) reported the gender of the caregiv-
ers. There were 20 males and 125 female caregivers. Of the 191 caregiver par-
ticipants, at least 132 (69.1%) were parents, 16 (8.4%) were school staff, 16 were
teachers, 11 (5.8%) were behavioral technicians, seven (3.7%) were undergradu-
ate students, five (2.6%) were graduate students, three (1.6%) were therapists, and
one was a teacher assistant. Ten (33.3%) of the 30 studies reported ages of the
caregivers. The average age of caregiver participants was 30 years, excluding one
study that reported the age range of seven caregivers as being from 21 to over
50 years. Eight studies (26.7%) reported the race/ethnicity of caregiver partici-
pants. Twenty-nine caregivers were Caucasian, three Hispanic, two Asian Ameri-
can, two multi-racial, one Middle Eastern, and one Latino. One study indicated
that all participants were from Saudi Arabia (Alnemary et al. 2015) and another
indicated that all participants were from the country of Georgia (Barkaia et al.
2017).
13
Study Design
Across the 30 studies, 23 (76.7%) used a single-case experimental design, five

(16.7%) used a group experimental design, and two did not use an experimen-
tal design. Of the studies using single case design, 12 used a multiple baseline
or multiple probe design, four used a multielement design, two used an ABAB
design, two used a multiple baseline with an embedded multielement design, one
used a multiple baseline with an alternating treatments design, one used a multi-
ple probe design with an embedded reversal, and one used an AB with a multi-
element design. Seven studies used an additional multielement design to present
structured descriptive assessment (SDA) and/or functional analysis (FA) data.
Of the studies using a group experimental design, three conducted a randomized
clinical trial involving random assignment to one of two intervention groups, one
conducted a randomized clinical trial with a waitlist control, and one conducted a
quasi-experiment with a pre- and posttest and no control group.
ABA‑Based Procedures
For each reviewed study, information was gathered on the types of behavioral pro-
cedures taught to caregiver participants. The procedures taught to caregiver partici-
pants most frequently in the studies were FA (n = 12), FCT (n = 9), and a combina-
tion of ABA-based teaching strategies (n = 6). Other procedures included preference
assessments (n = 4), SDAs (n = 4), the Early Start Denver Model (ESDM) (n = 3),
differential reinforcement procedures (n = 2), antecedent-based procedures (n = 2),
discrete-trial training (DTT) (n = 1), consequence-based strategies (n = 1), mand
and echoics training (n = 1), the good behavior game (GBG) (n = 1), most-to-least
prompting (n = 1), and reciprocal imitation training (n = 1).
Dependent Measures
Information was also gathered on the dependent measures targeted across the
reviewed studies. The focus of dependent measures for all studies was either behav-
ior exhibited by the child participants (n = 11), behavior exhibited by the caregiver
participants (n = 7), or both (n = 12). A variety of child responses were measured,
with the most common being problem behavior (n = 13). Additional child responses
measured that were less common included mands and/or echoics (n = 5), task com-
pletion (n = 3), imitation (n = 3), augmentative and alternative communication
(AAC) requests (n = 2), idiosyncratic behavior (n = 2), joint attention (n = 2), use
of language targets (n = 1), social communication initiations and responses (n = 1),
functional vocalizations (n = 1), spontaneous communication (n = 1), spontaneous
vocalizations (n = 1), choice making (n = 1), engagement (n = 1), and compliance
(n = 1). Caregiver responses measured included the procedural fidelity with which
behavioral procedures were implemented (n = 16), occurrences of the use of proce-
dures (n = 2), and quality and rate of implementing procedures (n = 1). One study
additionally measured levels of parent engagement with their child (Vismara et al.
13
2013) and another study additionally measured levels of parent–child interaction

(Vismara et al. 2012).
Outcomes
Outcomes of the participants’ performance were recorded for all studies. In 14 stud-
ies, all child participants exhibited positive outcomes following implementation of
ABA-based interventions, with the positive outcomes being a decrease in problem
behavior and/or an increase in skill acquisition. Four studies had mixed results in
that not all participants exhibited improved performance during intervention com-
pared to baseline. There were no studies that involved negative outcomes for any
child participants.
In 15 studies, all caregiver participants exhibited improved performance in imple-
menting ABA-based procedures while being coached by therapists. One study had
mixed results in which not all participants exhibited improved performance during
intervention compared to baseline. One study reported that one caregiver exhibited a
decrease in performance in one out of seven ABA-based teaching procedures during
post-training. All 11 studies that involved the use of FAs to assess child participants
indicated successful implementation from caregivers and identified functions for all
or most child participants. One study indicated the ability of teachers to conduct a
preference assessment with 100% fidelity and to identify preferences of the child
participants (Machalicek et al. 2009b).
Training Components and Session Type
Table 2 provides a list of training components described among the included studies
with each component defined. Figure 2 displays a bar graph depicting the training
components reported on the x-axis and the percentage of studies that used each of
the components on the y-axis. All included studies provided a description of training
components used to teach caregiver participants behavioral procedures. A variety
of training components were used across the studies. All studies used at least one
component to teach one type of procedure, with some studies including more than
one set of components for the various assessments and interventions implemented,
such as FAs and FCT. For example, in a study by Simacek et al. (2017), researchers
conducted within-session instruction to teach parents how to implement an SDA and
FA. They also provided written instructions prior to starting FA sessions. To teach
parents to conduct FCT and most-to-least prompting, researchers provided written
instructions and conducted within-session instruction with modeling and perfor-
mance feedback. Across studies, the most common procedures used during training
were: providing performance feedback to caregivers on their implementation of pro-
cedures (n = 26), within-session instruction (n = 25), and modeling (demonstration
of procedures provided by at least two therapists live or via a video model) (n = 18).
Other less common components included pre-session instruction (n = 15), written
instructions (n = 12), prompting (n = 11), online instruction (n = 10), collaborative
problem solving (n = 9), and practice outside of sessions (n = 7). Some components
13
Table 2 List of training components and definitions

Training components Definitions
Performance feedback Comments provided vocally or electronically to caregiver

participants on their performance in implementing behavio-
ral procedures
Within-session instruction Response-guided and individualized directions provided
vocally to caregiver participants during live sessions on
implementing a behavioral procedure
Modeling A demonstration provided by researchers either live or via a
video model to caregiver participants on how to implement
behavioral procedures
Pre-session instruction Vocal instruction provided to caregiver participants prior
to the start of the first assessment or intervention session
involving a presentation or review of content related to the
study, behavioral principles, behavioral procedures, the
rationale for implementing specific behavioral procedures,
and/or how behavioral procedures should be used in the
natural environment
Address questions A researcher designates time to answer questions caregiver
participants have on behavioral principles or procedures
Prompting a researcher prompts caregiver participants to implement
specific steps of a behavioral procedure as caregivers are
implementing it with a child participant or confederate
Written instructions Response-guided and individualized directions provided to
caregiver participants on how to implement behavioral
procedures; typically provided via an electronic document
or email
Online instruction Slideshows, lectures, and/or modules provided to caregiver
participants in an electronic format and involves content
related to behavioral principles and procedures; may
include a rationale for implementing specific behavioral
procedures and recommendations for how to implement
procedures in the natural environment
Training until meeting criteria Training or intervention continues until caregiver participants
meet a certain performance criterion in implementing
behavioral procedures
Collaborative problem solving Discussion between a researcher and caregiver participant
in which they address potential or current issues related to
implementing behavioral procedures with child partici-
pants, discuss how the procedures should be used outside
of live sessions, and/or develop target skills for caregiver or
child participants
Written manual/handout A written or typed guide provided to caregiver participants
with content related to behavioral principles and/or proce-
dures
Learning checks Exercises or quizzes assessing a caregiver participant’s
comprehension of behavioral principles and/or procedures;
does not involve an assessment of a caregiver’s skill level in
implementing behavioral procedures
Caregiver reflections Report from caregiver participants describing their previous
use of behavioral procedures with child participants
13
Table 2 (continued)
Training components Definitions
Practice outside of sessions Caregiver participants practice implementing behavioral

procedures with child participants outside of live sessions
Additional resources Websites or readings researchers provided to caregiver
participants on topics such as autism, behavioral principles,
implementing behavioral procedures, or other related topics
Review of data A researcher presents a caregiver participant with assessment
or intervention data and explains the results
Practice exercises A caregiver participant practices implementing behavioral
procedures with a child participant or confederate during
live sessions with a researcher
Booster training Any repeated training provided to caregiver participants due
to them not meeting a performance criterion in implement-
ing behavioral procedures
Role-play Caregiver participants either play the role of a child or inter-
ventionist to practice implementing behavioral procedures
together without the presence of a child or confederate
Video self-modeling Caregiver participants can observe their past performance
in implementing a behavioral procedure for the purpose of
learning correct/incorrect implementation of the proce-
dures; video may include embedded performance feedback
Caregiver assessment A test conducted by researchers on caregiver participant
progress/current skill level in implementing a behavioral
procedure
Homework assignments Assignments related to behavioral principles or procedures
given to caregiver participants to complete outside of live
sessions
Performance feedback on learning checks Comments given to caregiver participants on written exer-
cises/quizzes they have completed assessing their under-
standing of ABA principles or procedures
Summary of content A review provided to caregiver participants on behavioral
principles or procedures used during live sessions; provided
after live sessions
Caregiver goal-setting Any instance in which caregiver participants set goals for
themselves in learning behavioral concepts or procedures
Graphing sessions Instruction provided to caregiver participants on how to graph
and interpret assessment or intervention data
found least often in the reviewed studies included a review of assessment and/or
intervention data (n = 6), video self-modeling (n = 4), homework assignments
(n = 3), caregiver goal-setting (n = 1), and a summary of content covered during live
sessions (n = 1).
Across studies, sessions in which caregiver participants implemented behavioral
procedures with child participants or confederates were either recorded by the car-
egiver participants and sent to researchers or were directly observed live by research-
ers as they communicated with the caregiver participants via telehealth. In both of
these circumstances, researchers recorded data on caregiver and/or child participant
13
Fig. 2 Data depicting the percentage of studies (n = 30) that used the training components listed on the
x-axis
performance. Across the 30 studies, 27 involved live sessions, one involved live or
recorded sessions for each participant, one involved both live and recorded sessions,
and one involved only recorded sessions.
Procedural Fidelity of Caregiver Participants
Interventionists’ procedural fidelity was recorded for each study that provided per-
tinent information. Twenty-four of the 30 studies (80%) reported interventionists’
procedural fidelity in implementing behavioral procedures that researchers or thera-
pists taught them via telehealth. Seventy-nine percent (n = 19) of studies reported
procedural fidelity in the form of a percentage of steps completed accurately, 16.7%
(n = 4) reported fidelity in the form of rating scale scores, and one study reported
fidelity using a measure of percentage of intervals. Of the 19 studies reporting pro-
cedural fidelity in the form of a percentage of steps completed accurately, 89.5%
(n = 17) reported procedural fidelity levels of 90% or higher for at least one caregiver
participant (see Table 1) and 5.3% (n = 1) of the studies reported procedural fidelity
levels of less than 70% for one caregiver participant.
Discussion
The purpose of the current literature review was to provide an update on the cur-
rent published empirical research on the use of telehealth for providing ABA-based
services to individuals with disabilities and their caregivers. The current review also
sought to gather information regarding training components used by researchers via
telehealth and caregivers’ procedural fidelity in implementing procedures. Overall,
13
the current research provides support for telehealth as a service delivery mecha-
nism for providing ABA-based services. All single-case-design studies involving
intervention for child participants (53.3% of all included studies) resulted in behav-
ior change in the desired direction for 73 child participants. All single-case-design
studies involving training caregiver participants to implement intervention (40%
of all included studies) resulted in behavior change in the desired direction for 55
caregiver participants. Additionally, all group-design studies resulted in behavior
change in a desired direction for both child and caregiver participant groups receiv-
ing intervention. Such outcomes provide support for the use of telehealth for deliv-
ering ABA-based services.
In examining training components used across studies and reported procedural
fidelity levels of caregiver participants implementing ABA-based procedures, some
possible considerations are worth noting. The authors gathered data from 19 stud-
ies that included both descriptions of training components and procedural fidelity
levels in the form of a percentage of correct caregiver responses. Across the 19 stud-
ies, 89.5% reported procedural fidelity levels of 90% or higher for at least one car-
egiver participant and only one study reported procedural fidelity levels of less than
70% for one caregiver participant. However, within and across training components,
both relatively high and relatively low procedural fidelity levels were reported. No
specific training component was associated with exclusively high or low procedural
fidelity, suggesting that procedural fidelity is not exclusively influenced by the train-
ing component(s). To illustrate, there are examples of individual studies in which
caregiver participants received the same training and yet exhibited different levels
of procedural fidelity. For example, in a study by Suess et al. (2014), researchers
used a package of training components, including performance feedback, pre-ses-
sion instruction, within-session instruction, and practice outside of sessions. The
same training components were used for all three participating caregivers and aver-
age procedural fidelity scores across caregivers were quite variable (ranging from 74
to 94.1%). Thus, it appears there are other factors that influence procedural fidelity
aside from the training components.
Several possible factors may influence the level of procedural fidelity in which
caregivers implement assessment or intervention procedures. For one, the proce-
dures themselves could influence procedural fidelity, whether it be the number of
steps that need to be completed or the difficulty level of the procedures. Fidelity
may also be influenced by idiosyncratic caregiver factors including but not limited
to educational background, experience, and competing responsibilities. Therefore,
it may be beneficial for behavior analysts to consider using systematic methods to
develop individualized training packages for caregivers and to make adjustments
to training procedures when needed, just as they do when developing intervention
packages for individuals receiving ABA-based intervention.
One interesting finding is the prevalence of children with ASD compared to
children with other disabilities who participated in studies involving telehealth.
Of the 157 child participants across 30 studies, just over half (55.4%) were diag-
nosed with ASD, whereas nearly half of child participants had other diagnoses. Such
data provide some support for the use of telehealth to provide behavioral services
to individuals with a variety of disabilities. It also emphasizes the importance of
13
the current literature review in expanding the inclusion criteria in order to include
individuals with disabilities other than ASD, because such studies also contribute
information supporting the use of telehealth. Additionally, participants who were
implementing behavioral assessments or interventions represented a diverse popula-
tion, with a variety of professions and different education backgrounds, the majority
of which were parents of child participants. Some studies also included participants
who resided in nations outside of the USA, such as Canada, Saudi Arabia, and the
country of Georgia (Alnemary et al. 2015; Barkaia et al. 2017; Wainer and Ingersoll
2015). Thus, the current telehealth literature provides emerging evidence towards
the use of telehealth in providing ABA-based services to people with different back-
grounds and who reside in various countries.
Limitations to the Review
The current review should be considered in light of some limitations. First, five stud-
ies that included participants who had experience using the specific behavioral pro-
cedures being taught to them were not included in the current review. Including such
studies may have provided useful information about the status of telehealth as a ser-
vice tool that is excluded in this review. However, such studies were not included in
order to gather information on training components used for novice users of ABA-
based procedures. Second, specific relations between procedural fidelity levels and
training components are unknown.
Implications for Future Research
Future research on telehealth should continue to test the effectiveness of using

various training components remotely for teaching caregivers to implement behav-
ioral procedures in order to provide further evidence of the utility of telehealth as
an established format for delivering ABA-based services. Additionally, researchers
should use statistical methods, such as a meta-analysis, or experimental methods to
determine effective and essential components in training individuals to use behav-
ioral procedures accurately (Neely et al. 2017) or to determine other factors that
may influence procedural fidelity. Future research should also test whether various
training components are more effective in teaching caregivers to implement behav-
ioral procedures designed to address problem behavior versus those focused on skill
acquisition. For example, one may test the effectiveness of teaching caregivers to
implement an FA compared to discrete-trial teaching using response-guided or indi-
vidualized components, such as within-session instruction to coach, versus non-indi-
vidualized components, such as a written manual. Overall, seeking to identify cur-
rently effective components in telehealth is an important endeavor. Taking steps to
perfect the use of telehealth may allow children with disabilities and their families to
receive services they would otherwise not obtain and may lead to long-term positive
outcomes for those families.
13
Funding This study was funded by grant number 1R21DC015021.
Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict of interest.
References
*Indicates studies included in the review
*Alnemary, F. M., Wallace, M., Symon, J. B. G., & Barry, L. M. (2015). Using international videoconfer-
encing to provide staff training on functional behavioral assessment. Behavioral Interventions, 30,
73–86.
*Barkaia, A., Stokes, T. F., & Mikiashvili, T. (2017). Intercontinental telehealth coaching of therapists to
improve verbalizations by children with autism. Journal of Applied Behavior Analysis, 50, 582–589.
Barretto, A., Wacker, D. P., Harding, J., Lee, J., & Berg, W. K. (2006). Using telemedicine to conduct
behavioral assessments. Journal of Applied Behavior Analysis, 39, 333–340.
*Benson, S. S., Dimian, A. F., Elmquist, M., Simacek, J., McComas, J. J., & Symons, F. J. (2018).
Coaching parents to assess and treat self-injurious behaviour via telehealth. Journal of Intellectual
Disability Research, 62, 1114–1123.
Boisvert, M., Lang, R., Andrianopoulos, M., & Boscardin, M.-L. (2010). Telepractice in the assessment
and treatment of individuals with autism spectrum disorders: A systematic review. Developmental
Neurorehabilitation, 13, 423–432.
Bulgren, J. A. (2002). The educational context and outcomes for high school students with disabilities:
The perceptions of parents of students with disabilities. Research Report. Retrieved from https://
files.eric.ed.gov/fulltext/ED469288.pdf.
*Dimian, A. F., Elmquist, M., Reichle, J., & Simacek, J. (2018). Teaching communicative responses with
a speech-generating device via telehealth coaching. Advances in Neurodevelopmental Disorders, 2,
86–99.
Ferguson, J., Craig, E. A., & Dounavi, K. (2019). Telehealth as a model for providing behaviour ana-
lytic interventions to individuals with Autism Spectrum Disorder: A systematic review. Journal of
Autism and Developmental Disorders, 49, 582–616.
*Fischer, A. J., Dart, E. H., Radley, K. C., Richardson, D., Clark, R., & Wimberly, J. (2017). An evalua-
tion of the effectiveness and acceptability of teleconsultation. Journal of Educational and Psycho-
logical Consultation, 27, 437–458.
*Fisher, W. W., Luczynski, K. C., Hood, S. A., Lesser, A. D., Machado, M. A., & Piazza, C. C. (2014).
Preliminary findings of a randomized clinical trial of a virtual training program for applied behavior
analysis technicians. Research in Autism Spectrum Disorders, 8, 1044–1054.
*Gibson, J. L., Pennington, R. C., Stenhoff, D. M., & Hopper, J. S. (2010). Using desktop videoconfer-
encing to deliver interventions to a preschool student with autism. Topics in Early Childhood Spe-
cial Education, 29, 214–225.
*Hay-Hansson, A. W., & Eldevik, S. (2013). Training discrete trials teaching skills using videoconfer-
ence. Research in Autism Spectrum Disorders, 7, 1300–1309.
*Heitzman-Powell, L. S., Buzhardt, J., Rusinko, L. C., & Miller, T. M. (2014). Formative evaluation of
an ABA outreach training program for parents of children with autism in remote areas. Focus on
Autism and Other Developmental Disabilities, 29, 23–38.
*Higgins, W. J., Luczynski, K. C., Carroll, R. A., Fisher, W. W., & Mudford, O. C. (2017). Evaluation of
a telehealth training package to remotely train staff to conduct a preference assessment. Journal of
Applied Behavior Analysis, 50, 238–251.
*Ingersoll, B., Wainer, A. L., Berger, N. I., Pickard, K. E., & Bonter, N. (2016). Comparison of a self-
directed and therapist-assisted telehealth parent mediated intervention for children with ASD: A
pilot RCT. Journal of Autism and Developmental Disorders, 46, 2275–2284.
13
*Knowles, C., Massar, M., Raulston, T.-J., & Machalicek, W. (2017). Telehealth consultation in a
self-contained classroom for behavior: A pilot study. Preventing School Failure, 61, 28–38.
Knutsen, J., Wolfe, A., Burke, B. L., Hepburn, S., Lindgren, S., & Coury, D. (2016). A systematic
review of telemedicine in autism spectrum disorders. Review Journal of Autism And Develop-
mental Disorders, 3, 330–344.
Lindgren, S., Wacker, D. P., Suess, A., Schieltz, K., Pelzel, K., Kopelman, T., et al. (2016). Telehealth
and autism: Treating challenging behavior at a lower cost. Pediatrics, 137(2), 167–175.
*Machalicek, W., Lequia, J., Pinkelman, S., Knowles, C., Raulston, T., Davis, T., et al. (2016). Behav-
ioral telehealth consultation with families of children with autism spectrum disorder. Behavioral
Interventions, 31, 223–250.
*Machalicek, W., O’Reilly, M., Chan, J. M., Lang, R., Mandy, R., Tonya, D., et al. (2009a). Using
videoconferencing to conduct functional analysis of challenging behavior and develop classroom
behavioral support plans for students with autism. Education and Training in Developmental
Disabilities, 44(207), 217.
*Machalicek, W., O’Reilly, M., Chan, J. M., Rispoli, M., Lang, R., Davis, T., et al. (2009b). Using
videoconferencing to support teachers to conduct preference assessments with students with
autism and developmental disabilities. Research in Autism Spectrum Disorders, 3, 32–41.
*Machalicek, W., O’Reilly, M. F., Rispoli, M., Davis, T., Lang, R., Franco, J. H., et al. (2010). Train-
ing teachers to assess the challenging behaviors of students with autism using video teleconfer-
encing. Education and Training in Autism and Developmental Disabilities, 45, 203–215.
*Martens, B. K., Baxter, E. L., McComas, J. J., Sallade, S. J., Kester, J. S., Caamano, M., Dimian, A.,
Simacek, J., & Pennington, B. (2019). Agreement between structured descriptive assessments
and functional analyses conducted over a telehealth system. Behavior Analysis: Research and
Practice, 19, 343–356.
*Meadan, H., Snodgrass, M. R., Meyer, L. E., Fisher, K. W., Chung, M. Y., & Halle, J. W. (2016).
Internet-based parent-implemented intervention for young children with autism: A pilot study.
Journal of Early Intervention, 38, 3–23.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The PRISMA Group. (2009). Preferred report-
ing items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med, 6,
e1000097. https://doi.org/10.1371/journal.pmed1000097.
Murphy, M. A., & Ruble, L. A. (2012). A comparative study of rurality and urbanicity on access to
and satisfaction with services for children with autism spectrum disorders. Rural Special Educa-
tion Quarterly, 31(3), 3–11.
Neely, L., Rispoli, M., Gerow, S., Hong, E. R., & Hagan-Burke, S. (2017). Fidelity outcomes for
autism-focused interventionists coached via telepractice: A systematic literature review. Jour-
nal of Developmental & Physical Disabilities, 29, 849–874. https://doi.org/10.1007/s1088
2-017-9550-4.
Puskin, D., Johnston, B., & Speedie S. (2006). Telemedicine, telehealth, and health information tech-
nology: An APA issue paper [.doc file]. Retrieved June 22, 2019, from https://www.who.int/goe/
policies/countr ies/usa_suppor t_tele.pdf.
Reichow, B., Volkmar, F. R., & Cicchetti, D. V. (2008). Development of the evaluative method for
evaluating and determining evidence- based practices in autism. Journal of Autism and Develop-
Salomone, E., Besterah, K., Derbyshire, K., McCloy, C., Hastings, R. P., Howlin, P., et al. (2014).
Emotional and behavioural problems in children and young people with autism spectrum disor-
der in specialist autism schools. Research in Autism Spectrum Disorders, 8, 661–668.
*Schieltz, K. M., Romani, P. W., Wacker, D. P., Suess, A. N., Huang, P., Berg, W. K., et al. (2018).
Single-case analysis to determine reasons for failure of behavioral treatment via telehealth.
Remedial and Special Education, 39(95), 105.
Sutherland, R., Trembath, D., & Roberts, J. (2018). Telehealth and autism: A systematic search and
review of the literature. International Journal of Speech-Language Pathology, 20, 324–336.
*Simacek, J., Dimian, A. F., & McComas, J. J. (2017). Communication intervention for young chil-
dren with severe neurodevelopmental disabilities via telehealth. Journal of Autism and Develop-
*Suess, A. N., Romani, P. W., Wacker, D. P., Dyson, S. M., Kuhle, J. L., Lee, J. F., et al. (2014). Eval-
uating the treatment fidelity of parents who conduct in-home functional communication training
with coaching via telehealth. Journal of Behavioral Education, 23, 34–59.
13
*Suess, A. N., Wacker, D. P., Schwartz, J. E., Lustig, N., & Detrick, J. (2016). Preliminary evidence on
the use of telehealth in an outpatient behavior clinic. Journal of Applied Behavior Analysis, 49,
686–692.
*Sump, L. A., Richman, D. M., Schaefer, A. M., Grubb, L. M., & Brewer, A. T. (2018). Telehealth and
in-person training outcomes for novice discrete trial training therapists. Journal of Applied Behavior
Analysis, 51, 466–481.
Tomlinson, S. R. L., Gore, N., & McGill, P. (2018). Training individuals to implement applied behavior
analytic procedures via telehealth: A systematic review of the literature. Journal of Behavioral Edu-
cation, 27, 172–222.
*Vismara, L. A., McCormick, Carolyn E. B., Wagner, A. L., Monlux, Katernia, Nadhan, Anna, & Young,
Gregory S. (2018). Telehealth parent training in the early start Denver model: Results from a rand-
omized controlled study. Focus on Autism and Other Developmental Disabilities, 33, 67–79.
*Vismara, L. A., McCormick, C., Young, G. S., Nadhan, A., & Monlux, K. (2013). Preliminary findings
of a telehealth approach to parent training in autism. Journal of Autism and Developmental Disor-
ders, 43, 2953–2969.
*Vismara, L. A., Young, G. S., & Rogers, S. J. (2012). Telehealth for expanding the reach of early autism
training to parents. Autism Research and Treatment, 2012, 12.
*Wacker, D. P., Lee, J. F., Dalmau, Y. C. P., Kopelman, T. G., Lindgren, S. D., Kuhle, J., et al. (2013a).
Conducting functional analyses of problem behavior via telehealth. Journal of Applied Behavior
Analysis, 46, 31–46.
*Wacker, D. P., Lee, J. F., Dalmau, Y. C. P., Kopelman, T. G., Lindgren, S. D., Kuhle, J., et al. (2013b).
Conducting functional communication training via telehealth to reduce the problem behavior of
young children with autism. Journal of Developmental and Physical Disabilities, 25, 35–48.
*Wainer, A. L., & Ingersoll, B. R. (2015). Increasing access to an ASD imitation intervention via a tel-
ehealth parent training program. Journal of Autism and Developmental Disorders, 45, 3877–3890.
*Wilczynski, S. M., Labrie, A., Baloski, A., Kaake, A., Marchi, N., & Zoder-Martell, K. (2017).
Web-based teacher training and coaching/feedback: A case study. Psychology in the Schools, 54,
433–445.
Wong, C., Odom, S. L., Hume, K. A., Cox, A. W., Fettig, A., Kucharczyk, Suzanne, et al. (2015). Evi-
dence-Based Practices for Children, Youth, and Young Adults with Autism Spectrum Disorder: A
Comprehensive Review. Journal of Autism and Developmental Disorders, 45, 1951–1966.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional affiliations.
13

Unholz-Bowden2020_Article_CaregiverTrainingViaTelehealth

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Unholz-Bowden2020_Article_CaregiverTrainingViaTelehealth

Uploaded by

Copyright:

Available Formats

Journal of Behavioral Education (2020) 29:246–281

Caregiver Training Via Telehealth on Behavioral Procedures:

Emily Unholz‑Bowden1 · Jennifer J. McComas1 · Kristen L. McMaster1 ·

Published online: 2 April 2020

Keywords Telehealth · Autism spectrum disorder · Developmental disability ·

design and the demonstration of a functional relation or treatment effect. Second,

Previous literature reviews have provided a synthesis on literature related to car-

Search Procedures and Search Terms

In March 2019, the authors conducted a systematic literature search. We used

Records identified through database

Records after duplicates removed

Records screened Records excluded

Full-text articles assessed Full-text articles excluded,

Fig. 1 PRISMA flowchart depicting the literature search process

Search Agreement Procedures

Inclusion Agreement Procedures

Data Extraction Agreement Procedures

criteria, (j) collabo-

higher tional resources, (k)

Asian American fidelity of procedures implemented the assess- written instructions,

actions following ESDM outside of sessions,

Canada and child exhibited increased levels resources (k)

All information described below can be found in detail in Table 1.

In total, across the 30 reviewed studies, 27 included children with disabilities as

Across the 30 studies, 23 (76.7%) used a single-case experimental design, five

2013) and another study additionally measured levels of parent–child interaction

Training Components and Session Type

Table 2 List of training components and definitions

Performance feedback Comments provided vocally or electronically to caregiver

Practice outside of sessions Caregiver participants practice implementing behavioral

Procedural Fidelity of Caregiver Participants

Limitations to the Review

Implications for Future Research

Future research on telehealth should continue to test the effectiveness of using

Funding This study was funded by grant number 1R21DC015021.

Compliance with Ethical Standards

*Indicates studies included in the review

You might also like