Slaton Et Al-2017-Journal of Applied Behavior Analysis

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2017, 50, 252–277 NUMBER 2 (SPRING)
INTERVIEW-INFORMED FUNCTIONAL ANALYSES: A COMPARISON

OF SYNTHESIZED AND ISOLATED COMPONENTS
JESSICA D. SLATON
WESTERN NEW ENGLAND UNIVERSITY, NASHOBA LEARNING GROUP
GREGORY P. HANLEY
WESTERN NEW ENGLAND UNIVERSITY
AND
KATHERINE J. RAFTERY
NASHOBA LEARNING GROUP
Hanley, Jin, Vanselow, and Hanratty (2014) described a functional analysis (FA) format that
relied on a synthesis of multiple contingencies described by caregivers during open-ended
interviews. These interview-informed synthesized contingency analyses (IISCA) provided
effective baselines from which to develop socially validated treatments, but the synthesis pre-
cluded a precise understanding of individual contingencies influencing problem behavior. We
conducted IISCAs and standard FAs (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982/
1994) for nine children with autism to evaluate the likelihood of differentiation given a
number of synthesized versus isolated variables. The IISCA was differentiated for all. The
standard FA was differentiated for four; this number increased to six when we included pre-
cursors in the standard FA. We then compared treatments based on sets of differentiated
analyses for four children. Treatment based on the IISCA was effective for all four; treat-
ments based on the standard FA were effective for two. The role of synthesis in analysis is
discussed.
Key words: autism, functional analysis, functional communication training, IISCA, problem
behavior, synthesized analysis
Functional analysis (FA) is the most widely individuals (Beavers, Iwata, & Lerman, 2013;
researched method for assessing problem Hanley, Iwata, & McCord, 2003). Because
behavior of individuals with developmental the FA allows for the detection of relevant
disabilities and is a vital component in establishing operations (EOs) that evoke the
developing effective treatments for these problem behavior as well as the reinforcers
that maintain it, treatments for problem
behavior that are developed from an FA are
This study was conducted in partial fulfillment of a Ph. more likely to be effective (Campbell, 2003)
D. in Behavior Analysis from Western New England Uni-
versity by the first author. We thank Rachel Thompson,
and less likely to rely on punishment
Jason Bourret, and Jessica Sassi for their feedback on ear- (Kahng, Iwata, & Lewin, 2002; Pelios, Mor-
lier versions of this manuscript. We would also like to ren, Tesch, & Axelrod, 1999). Information
thank Matthew Mosher, Angie Bird, Megan Chambers, obtained from an FA may then be integrated
Kerin Griswould, Leanne Patenaude, Kendra Penny,
Rebecca Losavio, Mia Morgan, Eileen Sauer, and Jessica into a treatment such as functional commu-
Torres for conducting sessions, scoring data, or otherwise nication training (FCT), in which the identi-
assisting with this project. fied reinforcers are delivered contingent on
Address correspondence to: Gregory P. Hanley, West- an appropriate communication response
ern New England University, 1215 Wilbraham Road,
Springfield, MA 01119. Email: ghanley@wne.edu (e.g., Carr & Durand, 1985; Fisher et al.,
doi: 10.1002/jaba.384 1993; Tiger, Hanley, & Bruzek, 2008).
© 2017 Society for the Experimental Analysis of Behavior
252
SYNTHESIZED AND ISOLATED COMPONENTS 253
In their seminal article, Iwata, Dorsey, Slifer, repeated measures of problem behavior across
Bauman, and Richman (1982/1994) described multiple test conditions (i.e., a standard FA).
an FA that used a multielement design to com- However, when limited assessment time is a
pare test conditions for each of the most com- concern, a brief FA with only one session per
monly suspected operant classes of problem condition may be conducted (e.g., Derby et al.,
behavior (automatic reinforcement, socially 1992; Northup et al., 1991). When high-risk
mediated positive reinforcement, socially behavior is a concern, a latency FA may be
mediated negative reinforcement). Reviews of conducted in which the session is terminated
the FA literature by Hanley et al. (2003) and after the first response (e.g., Thomason-Sassi,
Beavers et al. (2013) indicate that most exam- Iwata, Neidert, & Roscoe, 2011) or a precursor
ples of published FAs share a number of fea- FA may be conducted in which reinforcement
tures in common with the analysis described by contingencies are applied to behaviors that
Iwata et al. (1982/1994). For example, Beavers tend to precede the high-risk behavior
et al. reported that 80% of FAs used a multiele- (e.g., Borrero & Borrero, 2008; Fritz, Iwata,
ment design, 90% had multiple test conditions, Hammond, & Bloom, 2013; Herscovitch,
over 90% used the ABC model in which conse- Roscoe, Libby, Bourret, & Ahearn, 2009;
quences are provided for target behaviors in Langdon, Carr, & Owen-DeSchryver, 2008;
each test condition, over 90% included a test Smith & Churchill, 2002). These formats are
for socially mediated positive reinforcement, presented as alternatives to the standard FA.
over 90% included a test for socially mediated A different type of concern, however, is how
negative reinforcement, and over 50% included to proceed when a standard FA has been con-
a test for automatic reinforcement. These per- ducted and is inconclusive. Modifications that
centages are similar to those reported by deviate more substantially from a standard FA
Hanley et al. (2003). Given that the majority may be necessary when inconclusive results are
of published FAs include these common fea- obtained, although it is unclear how often this
tures, it seems reasonable to describe an analysis may occur. As one example, Hagopian, Rooker,
with these features as a standard FA. This does Jessel, and DeLeon (2013) reported the out-
not mean that FA as a methodology is or comes of 176 FAs that were selected using “a
should be standardized; in fact, FA is a power- consecutive case-series design … to minimize
ful and flexible tool that can be adapted to fit a any potential selection bias favoring particular
variety of research or clinical needs (Hanley, outcomes” (p. 91). Of the standard FAs that
2012; Iwata & Dozier, 2008). However, the met their inclusion criteria, 53% (94 of 176)
term standard FA is a convenient way to tact were undifferentiated on the first attempt.
the particular type of FA that is most frequently These results should be interpreted with cau-
reported in the literature, and we use this term tion because they were drawn from a popula-
to refer to such FAs throughout the rest of the tion of individuals on an inpatient unit for the
article. treatment of severe problem behavior, and such
Iwata and Dozier (2008) summarized a a sample may not be representative of the larger
number of modifications to standard FA proce- population of individuals who engage in prob-
dures that have been reported over the past few lem behavior. However, these data are impor-
decades, each of which they described as best tant because the authors reported the number
suited to a particular set of circumstances. of iterations required before an FA was differ-
When there are very few constraints to the entiated. Other large n studies (e.g., the epide-
assessment process, these authors recommend miological study by Iwata, Pace, et al., 1994)
what they term a “full FA,” characterized by report only the final iteration of the FA,
254 JESSICA D. SLATON et al.
making it unclear the extent to which a stand- and the successful modification of the standard
ard FA is likely to be differentiated on the first FA with the inclusion of idiosyncratic variables.
attempt. The data reported by Hagopian Hanley et al. (2014) reported a set of sub-
et al. suggest that a standard FA may often stantial modifications to the standard FA that
need to be modified to obtain conclusive included many of the variations described
results. Continuing to evaluate FA modifica- above. These authors used an open-ended
tions and their likelihood of differentiation is interview (see Hanley, 2012) to identify rein-
therefore an important line of inquiry. forcers suspected of influencing problem behav-
One such possible modification is to com- ior, and these reinforcers were then synthesized
bine idiosyncratic reinforcers in a single test in one test condition that emulated a naturally
condition. For example, Bowman, Fisher, occurring context in which problem behavior
Thompson, and Piazza (1997) reported an FA was reported to occur. This synthesized test
with a single test condition in which the parti- condition was compared to a matched control
cipant’s mands were reinforced contingent on condition in which all putative reinforcers were
destructive behavior. In a second example, continuously available. This type of analysis has
Fisher, Adelinis, Thompson, Worsdell, and since been described as an interview-informed
Zarcone (1998) described an FA with a single synthesized contingency analysis or IISCA
test condition in which problem behavior pro- (Ghaemmaghami, Hanley, & Jessel, 2016;
duced escape from the analyst’s prompts and Ghaemmaghami, Hanley, Jin, & Vanselow,
interruptions plus access to resuming the ongo- 2015; Jessel, Hanley, & Ghaemmaghami,
ing activity in which the participant had been 2016). The IISCA packages multiple FA modi-
engaged. In a recent review, Schlichenmeyer, fications that have been reported separately by
Roscoe, Rooker, Wheeler, and Dube (2013) several authors: (a) a single test condition rather
described idiosyncratic variables that have been than multiple test conditions (e.g., Adelinis &
evaluated in FAs. These authors identified Hagopian, 1999; Bowman et al., 1997; Fisher
42 studies published between 2001 and 2010 et al., 1998); (b) combined rather than isolated
in which the standard FA was modified to reinforcers (e.g., Kuhn, Hardesty, & Sweeney,
include participant-specific rather than generic 2009; Leon, Lazarchick, Rooker, & DeLeon,
variables. Some examples of idiosyncratic EOs 2013; Mann & Mueller, 2009; Mueller,
included manipulating the type of prompting Sterling-Turner, & Moore, 2005; Payne,
for a task (Tiger, Fisher, Toussaint, & Kodak, Dozier, Neidert, Jowett, & Newquist, 2014;
2009), having the analyst engage with the parti- Rispoli, Camargo, Machalicek, Lang, &
cipant’s preferred items (Kuhn, Hardesty, & Sigafoos, 2014; Sarno et al., 2011; Zarcone,
Luczynski, 2009), or manipulating the level of Fisher, & Piazza, 1996); (c) idiosyncratic rather
social attention (Call, Wacker, Ringdahl, than generic EOs and reinforcers (see Schlichen-
Cooper-Brown, & Boelter, 2004). Some exam- meyer et al., 2013, for a review); (d) shorter ses-
ples of idiosyncratic reinforcers included a spe- sion duration (Wallace & Iwata, 1999);
cific type of attention (Kodak, Northup, & (e) reinforcing co-occurring topographies of
Kelley, 2007), access to ritualistic behaviors problem behavior (e.g., Lalli, Mace, Wohn, &
(Falcomata, Roane, Feeney, & Stephenson, Livezey, 1995; Magee & Ellis, 2000; Richman,
2010), or engaging in preferred conversations Wacker, Asmus, Casey, & Andelman, 1999);
(Roscoe, Kindle, & Pence, 2010). The exam- and (f ) reinforcing precursors to problem
ples reviewed by Schlichenmeyer et al. are behavior (e.g., Borrero & Borrero, 2008; Lang-
important because they demonstrate the value don et al., 2008; Smith & Churchill, 2002).
of considering individualized variables in FAs None of the components included in the IISCA
are novel in and of themselves; each has prece- identifying the individual impact of each rein-
dent in the FA literature. The combination of forcer suspected of influencing a particular
these particular modifications is novel and problem behavior. Combining variables that
has produced highly effective treatments are reported to occur together can offer preci-
(Ghaemmaghammi et al., 2016; Ghaemma- sion in emulating the natural contexts in which
ghami et al., 2015; Hanley et al., 2014; Santi- problem behavior typically occurs. At present,
ago, Hanley, Moore, & Jin, 2016; Strohmeier, we do not know the relative advantage of these
Murphy, & O’Connor, 2016). However, two types of precision in analyzing and treating
because multiple variables are synthesized in the problem behavior. For example, are we more
IISCA, the relevance of each variable in any likely to obtain a differentiated analysis and
demonstrated functional relation is unclear. develop an effective treatment by precisely iso-
The IISCA synthesizes each component of lating each response–reinforcer relation, or by
the three-term contingency: antecedents (EOs), precisely replicating the contexts under which
behaviors (response topographies), and conse- problem behavior occurs in the natural envi-
quences. Establishing operations are synthesized ronment? One way to begin answering these
by presenting multiple evocative events at once, questions may be to compare outcomes from
such as restricting access to tangible items while IISCAs to outcomes of standard FAs to deter-
presenting demands. Response topographies are mine whether both analyses are differentiated,
synthesized by including precursor responses whether they suggest different reinforcers for
and co-occurring topographies of problem problem behavior, and whether subsequent
behavior in the contingency class. For example, treatments based on both analyses are effective.
if foot stomping is reported to occur shortly Fisher, Greer, Romani, Zangrillo, and Owen
before bouts of self-injury and aggression, any (2016) reported data for five participants with
of those three responses would be reinforced whom they conducted an IISCA and standard
during the IISCA test condition. The inclusion FA. Both analyses were differentiated for four
of precursors has been reported by other out of five participants; the fifth participant did
authors who have conducted an IISCA (see Jes- not engage in any problem behavior in either
sel et al., 2016) and is one of its characteristic analysis. For cases in which both analyses were
features. Consequences are synthesized based on differentiated, the standard FA detected one or
reports of outcomes that tend to occur together two of the same reinforcers that had been
following problem behavior. For example, for synthesized in the IISCA. For example, for par-
participant Gail in Hanley et al. (2014), Gail’s ticipants Allie and Tina, the IISCA combined
mother reported providing her with preferred the reinforcers of escape, attention, and tangi-
toys and undivided attention to redirect the ble items; their standard FAs were differen-
problem behaviors of screaming, throwing tiated for tangible items. The data reported by
items, and aggression. Gail’s test session there- Fisher et al. (2016) indicate that the IISCA and
fore involved the provision of both tangible standard FA are both likely to be differentiated
items and high-quality attention for any and likely to detect some (though not all) of
instance of screaming, throwing items, or the same reinforcers for problem behavior.
aggression. Additional within-participant comparisons of
Synthesizing or isolating different compo- the IISCA and standard FA are needed for sev-
nents of contingencies offers different types of eral reasons. First, Fisher et al. (2016) repre-
precision in an FA. Providing a single reinforcer sents the only comparative evaluation of the
for a single topography of problem behavior in IISCA and standard FA thus far, and it is there-
each test condition can offer precision in fore difficult to speak to the generality of the
outcomes they obtained. Second, a comparative tangible items (when suggested by caregiver
treatment analysis was not included in that interview), escape, and automatic reinforce-
study, which means there are currently no data ment, and a generic toy-play control condition.
available regarding the relative efficacy of treat- The contingency in each test condition was
ments based on the IISCA and other FA for- applied to co-occurring topographies of prob-
mats. Because the primary purpose of an FA, lem behavior but not to precursors, as Beavers
especially in practice, is to obtain information et al., Hanley et al., and Iwata and Dozier all
that can be used to develop an effective treat- describe a precursor FA as a modification to
ment for problem behavior, it is particularly the standard FA rather than a characteristic fea-
important to know whether the IISCA yields ture of it. It should be noted, however, that if
treatments that are equally as effective as treat- precursor responses and problem behavior
ments based on other FA formats. We therefore occurred simultaneously or within seconds of
compared outcomes of IISCAs to outcomes of each other (e.g., whining while engaging in
standard FAs for nine children with autism SIB), the precursor response contacted rein-
(Study 1), and then compared outcomes of forcement by virtue of its close temporal prox-
IISCA-based treatment to outcomes of standard imity to problem behavior. If the only
FA-based treatment for a subset of those same responses observed during the standard FA
children (Study 2). were precursors (i.e., no problem behavior
occurred), we then conducted a precursor FA
as a third analysis. In summary, Study 1 com-
STUDY 1: COMPARISON OF IISCA pared the IISCA package as it has been most
AND STANDARD FA recently described to the standard FA as it has
The purpose of Study 1 was to compare the been most frequently described, with the use
IISCA and the standard FA to determine rela- of one modification to the standard FA
tive probability of differentiated outcomes. We (a precursor FA) when applicable. Variables
conducted this comparison with nine children such as session duration, analyst, specific tangi-
with autism, and we implemented each analysis ble items, specific demands, and specific types
format the way each has been reported most of attention were held constant across both
often in the literature. Based on descriptions analyses.
provided by Hanley et al. (2014) and Jessel Participants experienced the IISCA first and
et al. (2016), the IISCA therefore included the then the standard FA. We chose not to coun-
following features: a single test condition and terbalance the order of FA formats across parti-
matched control condition developed from an cipants because this tactic would not necessarily
open-ended interview, a synthesis of idiosyn- eliminate sequence effects (see Higgins Hains &
cratic EOs in the test condition, a synthesis of Baer, 1989). Instead, the order of IISCA first
idiosyncratic reinforcers provided contingent and standard FA second was arranged to capi-
on precursor or problem behavior in the test talize on a potential sequence effect: exposure
condition, and the use of a multielement design to some of the contingencies in the first analysis
to compare test and control conditions. Based could make it more likely that the second anal-
on methods described by Iwata et al. (1982/ ysis may be differentiated. We chose to give
1994) and data and practice recommendations this potential advantage to the standard FA. In
provided by Beavers et al. (2013), Hanley addition, when evaluating new FA formats,
et al. (2003), and Iwata and Dozier (2008), the there is precedent for conducting the new for-
standard FA included a multielement design mat first and then comparing results to those
with separate test conditions for attention, from a standard FA (e.g., Bloom, Iwata, Fritz,
Roscoe, & Carreau, 2011; Langdon et al., significantly restricted involvement in educa-
2008; Thomason-Sassi et al., 2011). tional and community activities for all partici-
After obtaining differentiation in the IISCA pants. All participants engaged in topographies
and then experiencing the same number of test of problem behavior that could produce bruises
sessions in the standard FA, there were four or abrasions to themselves or others; seven par-
possibilities for the participant: (a) if problem ticipants had treatment plans that included
behavior occurred in the standard FA and was emergency measures such as protective equip-
differentiated, participation in Study 1 was ment, physical restraint, or exclusionary time-
complete; (b) if problem behavior occurred in out; four participants were unable to attend
the standard FA but was not yet differentiated, school field trips or outings because of the
we conducted one or two more series of condi- chance that problem behavior may occur in the
tions to determine if responding would differ- community; four participants were taking med-
entiate with further analysis; (c) if no problem ications to address problem behavior; one par-
behavior occurred in the standard FA but pre- ticipant was often an hour late to school
cursors did occur, we conducted a precursor FA because of problem behavior during his morn-
to determine the effects of the standard FA ing routine at home.
contingencies on precursor responses that had All participants had a primary diagnosis of
been reinforced in the IISCA; and (d) if no autism; one participant had an additional diag-
problem behavior or precursors occurred in the nosis of Landau-Kleffner Syndrome (LKS), and
standard FA, we replicated the IISCA to verify one participant had an additional diagnosis of a
that the functional relation it had initially tic disorder. Participants ranged in age from
detected was still operating. 7 to 18 years old, and included seven boys and
two girls. The primary mode of communica-
tion for participants included vocal speech or
Method the use of speech-generating devices (SGDs)
Participants and setting. Participants were such as iPads™. Participants were reported to
children who attended a private day school for engage in a variety of problem behaviors:
individuals with autism. Clinical teams at the aggression, SIB, disruption, eloping, flopping,
school were informed that the first author climbing on furniture, and screaming. See
would be conducting a study on different FA Table 1 for more details regarding participant
formats and were asked if they had any clients characteristics.
for whom they would like assistance in treating Sessions were conducted in a 3.5 x 3.5 m
problem behavior. Participants were therefore room (Chloe, Diego, Dylan, Jeff, Jonah, Kyle,
nominated by their clinical teams based on the and Riley) or the participant’s classroom (Emily,
need for treatment of problem behavior, rather Mason). Session rooms were equipped with a
than selected by the authors. The participants table, chairs, a bookshelf, a small beanbag, and a
in Study 1 represent the first nine children soft mat. Participant-specific materials were
referred to the study; this study therefore con- added as applicable (e.g., specific toys, books,
forms to the requirements of a consecutive con- electronics). Classrooms were approximately 6 x
trolled case series design (Hagopian et al., 6 m and included individual workspaces for
2013). These participants either had a history three or four students, with each workspace
of failed analysis or treatment attempts, or had consisting of a desk and two chairs, small book-
recently displayed some change in problem shelves, and drawers with teaching materials.
behavior such as new topographies or an Response definitions, measurement, and inter-
increase in frequency. Problem behavior observer agreement (IOA). Sessions were
Table 1
Participant Characteristics
Participant Age Sex Diagnosis Language Abilitya Communication Mode Problem Behavior
Diego 11 M Autism 3 Vocal Aggression, SIB, disruption
Mason 10 M Autism 4 Vocal Aggression, disruption
Riley 12 M Autism 3 Vocal, SGD Aggression, SIB, disruption
Kyle 17 M Autism, tic disorder 3 SGD Aggression, SIB
Jonah 7 M Autism 2 SGD Eloping, flopping and hiding, climbing
Emily 13 F Autism 4 Vocal Aggression, SIB, screaming
Jeff 18 M Autism, LKS 2 Vocal, SGD Aggression, SIB, disruption, flopping
Dylan 7 M Autism 3 Vocal Aggression, disruption
Chloe 14 F Autism 3 Vocal Aggression, SIB
Note. SGD = speech generating device; SIB = self-injurious behavior; LKS = Landau-Kleffner syndrome
a
1 = no independent communication; 2 = single-word utterances; 3 = short sentences; 4 = full fluency
recorded using a camcorder or similar equip- engaging in aggression) and were defined indi-
ment, and data were collected using laptop vidually by participant. A response was
computers with data collection and analysis included as a precursor if the team reported
software. We scored each instance of problem that it tended to occur right before problem
behavior and precursor behavior, as well as the behavior and did not tend to occur at other
onset and offset of each reinforcement interval. times in the absence of problem behavior. See
From these data, we calculated rate of problem Table 2 for a list of the specific problem and
behavior, rate of precursors, and the condi- precursor behaviors reported for each
tional probability of a precursor response con- participant.
tacting reinforcement within 2 s during the IOA was assessed by having a second
standard FA. We report problem behavior and observer independently score at least 20% of
precursors as responses per min. In general, sessions (range, 20%–67%) in each condition
problem behavior consisted of the specific beha- of each analysis for each participant. Agreement
viors for which the participant had been was calculated by dividing sessions into 10-s
referred (e.g., aggression, SIB, disruption). intervals and dividing the number of agree-
Topographies of aggression included grabbing, ments per interval by the number of disagree-
pinching, scratching, hitting, biting, kicking, ments plus agreements per interval and
pulling hair, and head butting. Topographies of multiplying by 100. Mean IOA was 97% (ses-
SIB included hand-to-head, head-to-object, and sion range, 80%–100%).
self-biting. Topographies of disruption included Experimental design. We used a multielement
throwing work materials or toys, throwing fur- design to compare conditions within each anal-
niture, or ripping and destroying materials. ysis. Sessions were 3 min (Chloe, Riley), 5 min
Flopping was defined as dropping to the floor (Diego, Dylan, Emily, Jonah, Kyle, Mason), or
from a standing position or from a seated posi- 10 min (Jeff ). Session duration was determined
tion; eloping was defined as running more than in advance based on staff input and was held
two steps away from the teacher (in an open constant across analyses. For example, staff
space) or pushing down the door handle and reported that Jeff very frequently worked for
attempting to leave the room (in a closed periods of 5 min or longer without problem
room). Precursors were responses that had been behavior, and that problem behavior was more
reported to precede problem behavior likely to occur as his work task approached
(e.g., grimacing and clenching teeth before 10 min in duration or when he was required to
Table 2
Topographies of Problem Behavior and Precursors
Participant Problem Behaviors (reported to occur together) Precursor(s)

Diego Hitting, kicking, or head butting others; throwing furniture and other Screaming
objects; head-to-object SIB
Mason Hitting or kicking others; hitting or throwing objects Whining; grimacing; body slam in chair
Riley Hitting, kicking, or biting others; pulling hair; hand-to-head SIB; throwing Whining; body tensing
objects
Kyle Grabbing others; hand-to-head SIB Whining; biting lip
Jonah Eloping; flopping and hiding (flop plus legs or shoulders under an object); None reported
climbing on furniture
Emily Grabbing, scratching, or charging at others; biting self; head-to-object SIB; Whining; clenching teeth; sudden body
screaming jerks
Jeff Grabbing or pushing others; biting self; ripping materials; flopping “Nooo!” in high-pitched voice
(kneeling or laying on ground)
Chloe Grabbing or pushing others; hand-to-head SIB Screaming
Dylan Hitting, kicking, or charging at others; throwing objects at others Yelling “ASHHH!”
do more than one task in a row. We therefore reported in the interview to emulate the combi-
set Jeff’s sessions at 10 min. Staff reported that nation of EOs and consequences that were also
Chloe immediately became agitated the reported. By contrast, the role of the interview
moment anyone attempted to help her put on in the standard FA was to select stimuli with
her shoes, which is a very short task. We there- which to populate otherwise generic test condi-
fore set Chloe’s sessions at 3 min. tions whose contingencies were already
determined.
Multiple informants were interviewed
Procedures because the structure of the school program
Open-ended interview. Open-ended inter- was such that different staff interacted with
views can be helpful in gathering information the participant in different contexts, and there-
that may then be more formally and systemati- fore may have had different information to
cally evaluated (e.g., Hanley et al., 2014; Haw- share. We conducted one interview with all
kins, 1979; Iwata, Wong, Riordan, Dorsey, & informants at the same time for the sake of
Lau, 1982). We interviewed each participant’s efficiency, and so informants could add to
Board Certified Behavior Analyst (BCBA) and each other’s reports. The smallest group of
at least one other staff member who worked informants for any participant was two; the
directly with the participant. Each interview largest group was five. We searched for discre-
was approximately 30 min and involved some pancies between informant reports not because
form of the questions from the open-ended we were concerned with inter-informant relia-
functional assessment interview provided in the bility, but because we wanted to include the
appendix of Hanley (2012). Additional clarify- most potentially evocative situations and rein-
ing questions were asked as necessary. Because forcing consequences that were relevant to the
the standard FA for each participant included participant. For example, if one person
the same tangible items, demands, and types of reported that math tasks tended to evoke
attention that were synthesized in the IISCA, problem behavior but another person insisted
both analyses were therefore informed by the that daily living tasks were most problematic,
interview to some extent. However, the inter- we would ask additional questions such as
view was more intimately tied to the IISCA in how quickly each task seemed to evoke prob-
that the IISCA used the particular stimuli lem behavior, which task seemed to evoke
more dangerous topographies, which task was test, test.2 We selected this sequence because
of greatest concern to the team and family, alternating between control and test conditions
and whether the discrepant reports may be for two sessions of each provided the minimum
occurring because the informants interact with data necessary to demonstrate a functional rela-
the participant in different contexts. Rather tion; adding a third test session then provided
than viewing these discrepancies as indicating three data points to be used as a baseline from
poor reliability of the interview, we viewed which to evaluate treatment. Contingencies in
them as reflective of the fact that problem the IISCA test condition were unique to each
behavior occurs in many different contexts, participant (i.e., the open-ended interviews did
and we used follow-up questions to narrow not yield the same suspected combination of
down which specific context may be most rel- stimuli for any two participants). Contingencies
evant and practical to include in the analysis. included escape from a handwriting task to toys
IISCA. The control condition for each par- and attention (Diego); escape from a reading
ticipant included continuous access to the mul- task to toys, preferred conversations, and
tiple putative reinforcers arranged in the test answers (Mason); escape from medicine presen-
condition, thereby eliminating any EO for tation to YouTube (Riley); escape from an
these reinforcers. Problem behavior or precur- unpredictable to a predictable schedule (Kyle);
sors in the control condition produced no pro- escape from independent leisure tasks to toys,
grammed stimulus change. It is important to attention, and stereotypy (Jonah); escape from
note that to be included in the analysis, puta- medical demands to toys (Dylan); escape from
tive reinforcers must have been reported in the reading and daily living tasks to music and
interview as sometimes following problem child-directed interaction (Emily); escape from
behavior. In other words, we carefully com- a dressing task to rituals (Chloe); escape from
bined potential reinforcers based on informant vocational work to YouTube (Jeff ). See
report. We also carefully excluded stimuli based Table 3 for a summary of the control and test
on informant report. For example, attention in conditions of each individual IISCA.
the form of praise is often delivered on a fixed Standard FA. The standard FA for each par-
schedule during FA control conditions. How- ticipant included the following conditions:
ever, if attention was not one of the reinforcers alone or ignore, attention, tangible (except
specifically being evaluated in the IISCA test Chloe), escape, and play. These conditions
condition, we did not include it in the control were similar to those described by Iwata
condition. We responded to any social bids et al. (1982/1994) and were implemented in a
that the participant initiated, but we did not fixed sequence as described by Iwata, Pace,
otherwise attend to him or her and never pro- et al. (1994) to capitalize on EOs present across
vided any unsolicited attention. The test condi- sessions (e.g., experiencing the alone condition
tion contained the multiple EOs reported to first may establish the value of attention as a
evoke problem behavior, and the occurrence of reinforcer). As many variables were held con-
problem behavior or precursors produced 30 s stant between the IISCA and standard FAs as
of access to the multiple putative reinforcers.1 possible: session duration, analyst, type of
With one exception, the sequence of conditions
for all participants was control, test, control,
2
In Jeff’s case, because only one instance of problem
behavior occurred during his first test session, we contin-
1
In other words, the 30-s reinforcement interval in each ued alternating for a total of three control and four test
participant’s test condition was identical to his or her con- sessions to confirm that a functional relation was indeed
trol condition. present.
Table 3
IISCA Test and Control Conditions
Participant Control Condition (and test condition reinforcement) Test Condition (establishing operation)
Diego Read out loud to Diego and provide access to preferred Stop reading to Diego; instruct him to put down toys and
toys, with no handwriting demands presented come do handwriting
Mason Give Mason his toys and respond to any conversation Remove toys and instruct Mason to come do a reading task.
topics he initiates; answer any questions he asks Present words to read and do not otherwise interact with
Mason.
Riley Allow Riley to watch YouTube; do not present medicine Pause YouTube and instruct Riley to take his medicine
Kyle Allow Kyle to follow his schedule of preferred activities; do Interrupt Kyle’s schedule of preferred activities and present a
not change or interrupt his schedule difficult task (counting money) that is not on his schedule
Jonah Let Jonah go to his “fun space” to run around and tap Instruct Jonah to leave his “fun space” (empty room where he
things; provide undivided attention and eye contact can run around and tap things) and come put together a
puzzle at his desk
Emily Turn on music and allow Emily to direct all interactions Turn off music, instruct Emily to brush hair or pack up her
(e.g. repeat phrases for her or re-enact mistakes as backpack while correcting and physically prompting her
directed). Do not speak unless Emily speaks first.
Jeff Allow Jeff free access to YouTube and do not present any Pause YouTube and instruct Jeff to engage in vocational
demands counting and packaging task
Chloe Allow Chloe to put on / remove shoes her own way Interrupt shoe ritual and help Chloe put on / remove shoes
without interrupting any ritualistic behavior
Dylan Allow Dylan free access to his toys and do not try to take Tell Dylan it’s time to stop playing with toys and get his
his temperature temperature taken
attention, specific tangible items, and specific part of the particpant’s synthesized conse-
demands. quence in the IISCA, the same type of atten-
tion was provided in the standard FA. If
• Alone/ ignore condition. The alone condition attention was not part of the participant’s
(Mason, Riley, Kyle, Emily, Chloe) involved synthesized contingency in the IISCA, the
the participant sitting alone in a room or analyst provided generic attention in the
cubicle after the analyst said, “I’ll be back in form of reprimands and statements of
a few minutes.” The ignore condition (all concern.
other participants) was identical to the alone • Tangible condition. A tangible condition was
condition except that the analyst sat in a cor- included if access to tangibles was part of the
ner of the room and did not respond to any participant’s synthesized contingencies in the
participant behavior. An ignore rather than IISCA. Prior to the start of the condition,
alone condition was used in cases in which participants were given 1 min access to the
the team reported that the participant would same tangibles that had been provided in the
probably attempt to leave the room if left IISCA. After 1 min, the tangible items were
alone. then removed and no other interaction with
• Attention condition. The analyst told the par- the analyst occurred. Contingent on problem
ticipant that she had some work to do, and behavior, the tangible items were returned to
then turned away from the participant (but the participant for 30 s. Attention was not
remained within reach). Contingent on provided during the 30 s reinforcement inter-
problem behavior, the analyst provided the val beyond what was required to approach
participant with 30 s of attention. A full the participant silently and give him/her the
30 s of attention was delivered so that the items, to ensure that tangible versus attention
reinforcement intervals for all sessions of all contingencies remained as distinct as possible
analyses would be equal (e.g., Fisher, (e.g., Moore, Mueller, Dubard, Roberts, &
Piazza, & Chiang, 1996). If attention was Sterling-Turner, 2002).
• Play condition. During this condition, the Conditional probability analysis. Several parti-
same items from the tangible condition were cipants engaged in precursor and problem
available to the participant and attention was behavior simultaneously or within a few sec-
provided on an FT-30 s schedule and when- onds of each other during the standard FA,
ever the participant initiated an interaction. making it likely that precursor responses were
This attention was provided in the form of actually reinforced in both analyses for these
comments, compliments, or praise. If an participants. The conditional probability of a
interaction was participant-initiated, the ana- precursor response contacting reinforcement
lyst provided the particular type of attention during the standard FA was calculated by
requested or from the interview. If a tangible reviewing the data streams and counting the
condition was not included in the FA, the number of precursor responses that were fol-
participant was provided with items that tea- lowed by reinforcement within 2 s and dividing
chers reported were preferred. These were by the total number of precursor responses that
provided because the play condition in the occurred. This yielded a decimal value ranging
standard FA typically includes access to toys from 0 (no precursor responses contacted rein-
(Hanley et al., 2003; Iwata et al., 1982/ forcement within 2 s) to 1 (all precursor
1994). Problem behavior produced no pro- responses contacted reinforcement within 2 s).
grammed stimulus change. This analysis was performed for all participants
• Escape condition. The same demands pre- who engaged in precursor responses during the
sented in the IISCA were presented during standard FA.
this condition. Contingent on problem
behavior, 30 s of escape was provided. The Results
analyst removed the task materials and told
The IISCA was differentiated for 9/9 or
the participant he/she did not have to do the
100% of participants, the standard FA was dif-
task right now. No other attention was pro-
ferentiated for 4/9 or 44% of participants, and
vided, and the participant was not allowed to
the precursor FAs for both Mason and Riley
continue interacting with the task materials.
were differentiated (see Figures 1, 2, and 3 for
Withholding attention and access to the task
individual participant results; see Table 4 for a
materials during the escape interval was done
summary). Mean duration for the IISCA was
to ensure that the escape contingency
28 min (range, 15–70 min; SD 16) and mean
remained separate from any attention and
duration for the standard FA was 90 min
tangible contingencies.
(range, 36–200 min; SD 49).3 Thus, when
Precursor FA. The precursor FA was identical
both formats were implemented the way they
to the standard FA described above, except that
have been most often described in the litera-
any precursors that had been reinforced in the
ture, the IISCA was differentiated more often
IISCA were also reinforced in this analysis, in
than the standard FA and took less time to
addition to problem behavior. This means that
conduct.
the only difference between the IISCA and the
An analysis was considered differentiated if
precursor FA was the synthesis or isolation of
rates of responding in one or more test condi-
consequences. A precursor FA was conducted
tions were elevated relative to rates in the con-
with any participant for whom precursors were
trol condition. Regarding precursors and
the only response topography observed in the
problem behavior, it is possible that both
standard FA (i.e., no problem behavior
occurred during the standard FA, but precur- 3
Analysis duration was calculated by multiplying session
sors did occur). duration in minutes by total number of sessions.
occurred during an analysis, that only precur- replication of the IISCA was differentiated for
sors occurred, or that only problem behavior Diego; a subsequent precursor FA was differen-
occurred. Regarding differentiation, we tiated for Mason and Riley. Mason’s precursor
observed four distinct patterns in the standard FA was differentiated for escape, which was one
FA: (a) the standard FA was undifferentiated of the contingencies included in his IISCA;
because no problem behavior occurred; (b) the Riley’s precursor FA was differentiated for
standard FA was undifferentiated because no escape and tangibles, which were the same two
problem behavior occurred, but a subsequent contingencies included in his IISCA.
precursor FA was differentiated; (c) the stand- Figure 2 shows the results for two partici-
ard FA was undifferentiated because problem pants (Kyle and Jonah) for whom problem
behavior occurred but was uncontrolled; behavior occurred during the standard FA, but
(d) the standard FA was differentiated. was ultimately undifferentiated. For Kyle, prob-
Figure 1 shows the results for three partici- lem behavior was not reliably evoked in any
pants for whom the standard FA was undiffer- particular standard test condition, although pre-
entiated because problem behavior did not cursors persisted unreinforced across all condi-
occur. Diego (top panel) displayed no respond- tions. For Jonah, problem behavior persisted
ing during the standard FA; Mason and Riley uncontrolled across all standard FA conditions.
(second and third panels, respectively) dis- Figure 3 shows the results of four partici-
played precursor responses only. A subsequent pants (Dylan, Emily, Chloe, Jeff ) for whom
IISCA Standard FA Third Analysis

Topographies
reinforced PC and PB PB only PC and PB
2 Escape to toys IISCA replication
PC PB
and attention Ignore/Alone
Attention
Tangible
1 Escape
PC PB
Test Play
Control Diego
Precursors (PC) and problem
0
behavior (PB) per min
3 Standard precursor FA
Escape to toys,
preferred conversation,
2 answers
1 Mason
3 25 Standard precursor FA
Escape to
YouTube 20
2
15 Riley
10
1
5
0 0
1 2 3 4 5 5 10 15 5 10 15 20 25
Sessions
Figure 1. IISCA, standard FA, and follow-up analyses for Diego, Mason, and Riley.
Figure 2. IISCA and standard FA results for Kyle and Jonah. The y-axis denoting precursors and problem behavior
per min refers to only problem behavior for Jonah, for whom no precursors were identified.
the IISCA and standard FA were both differen- addition, there were three participants (Diego,
tiated. Dylan’s IISCA was differentiated for Jonah, Chloe) who did not engage in any precur-
escape to toys; his standard FA was differen- sors during either analysis, and instead engaged
tiated for escape and toys separately. Emily’s in problem behavior exclusively. Thus, there
IISCA was differentiated for escape to music were five total participants for whom identical
(tangibles) and child-directed interaction (atten- response topographies were reinforced in both
tion); her standard FA was differentiated for analyses. Of the remaining four participants, con-
escape only. Chloe’s IISCA was differentiated ditional probability values for each session ranged
for escape to rituals, and her standard FA was from 0.0 to 0.5 for Jeff and Kyle (averages 0.08
differentiated for escape only. It is important to and 0.07, respectively), indicating that some pre-
note, however, that we did not include an cursors were followed by reinforcement during
access to rituals condition in her standard FA some sessions, but most were not. Conditional
because this type of idiosyncratic reinforcer is probability values were 0 for Mason and Riley,
not typically evaluated in a standard FA. Jeff’s indicating that no precursor responses were ever
IISCA was differentiated for escape to You- followed by reinforcement during the standard
Tube, and his standard FA was differentiated FA (because no problem behavior occurred and
for attention. therefore reinforcement was never delivered dur-
There were two participants (Emily, Dylan) ing the standard FA).
for whom conditional probability values for each For the four participants for whom both ana-
session ranged from 0.8 to 1.0 (averages 0.9 and lyses were differentiated, without additional
0.96, respectively), indicating that almost all pre- data we could not determine which analysis
cursors in the standard FA were reinforced. In (if either) had accurately identified the
IISCA Standard FA
Topographies
PC and PB PB only
reinforced
5 PC PB PC PB
Test Ignore/Alone
4 Control Attention
Tangible
3 Escape to Escape
toys Play
2
1 Dylan
0
Precursors (PC) and problem behavior (PB) per min
4 Escape to
music and
3 child-directed
interaction
2
1 Emily
3
Escape to rituals
1
Chloe
2
Escape to
YouTube
1
Jeff
0
1 2 3 4 5 6 7 5 10 15 20
Sessions
Figure 3. IISCA and standard FA results for Dylan, Emily, Chloe, and Jeff.
reinforcers maintaining problem behavior, Iwata, 2007) and it may seem reasonable to
because both analyses suggested different func- apply this precedent as a means for interpreting
tional relations. Although there is precedent for IISCA results, the same logic does not hold
validating the results of descriptive assessments when comparing two FAs to each other. Both
by comparing them to the results of a standard the IISCA and standard FA are experimental;
FA (e.g., Camp, Iwata, Hammond, & Bloom, both are capable of demonstrating a functional
2009; Hall, 2005; Iwata, DeLeon, & Roscoe, relation. We therefore conducted a second
2013; Lerman & Iwata, 1993; Mace & Lalli, study to evaluate relative treatment efficacy. In
1991; St Peter et al., 2005, Thompson & Study 2, we compared function-based
Table 4
Functional Analysis Outcomes
Participant Function via IISCA Function via Standard Function via Standard Analysis of Precursors
Diego Escape to toys, attention Undifferentiated –
Mason Escape to toys, preferred conversation, answers Undifferentiated Escape
Riley Escape to YouTube Undifferentiated Escape, YouTube
Kyle Escape to predictable schedule Undifferentiated –
Jonah Escape to toys, attention, stereotypy Undifferentiated –
Emily Escape to music, child-directed time Escape –
Jeff Escape to YouTube Attention –
Chloe Escape to rituals Escape –
Dylan Escape to toys Escape, toys –
treatments developed from the IISCA and the same data collection and analysis program.
standard FA to determine which treatment was Response definitions and measurement for
more effective in eliminating problem behavior problem behavior and precursors were identical
and establishing an appropriate communication to Study 1. In addition, we scored each inde-
response for these four participants. pendent occurrence of the FCR. An occurrence
of the FCR was considered independent if the
participant began emitting the phrase within
STUDY 2: COMPARISON OF
5 s of the evocative event and without any
FUNCTION-BASED TREATMENTS
prompts from the analyst. We report problem
We directly compared IISCA-based and behavior (including precursors) and independ-
standard FA-based functional communication ent FCRs as responses per min. Data from each
training (FCT) plus extinction (EXT) treat- participant’s IISCA test condition and their dif-
ments for problem behavior for the four partici- ferentiated standard test condition(s) are
pants in Study 1 whose standard FAs were reported as baseline data against which to judge
differentiated. We evaluated each treatment treatment effects. Procedures for calculating
according to the outcomes of (a) eliminating IOA were identical to Study 1. IOA averaged
problem behavior and (b) generating an appro- 93% (session range, 87%–97%).
priate functional communication Experimental design. We used a multielement
response (FCR). design to compare the effects of FCT + EXT
based on each analysis. All sessions were
Method 10 trials or 10 min (whichever came first).
Participants and setting. Emily, Jeff, Chloe, Sessions were conducted 2 to 3 days per week,
and Dylan participated in this study. All treat- in blocks of two to six sessions (when compar-
ment sessions were conducted by the same ana- ing two FCT conditions) or blocks of three
lyst who conducted their FAs. The same sessions (when comparing three FCT condi-
stimuli that had been included in the relevant tions). Sessions were conducted in a random
FA condition were also used in treatment order each day, with the requirement that each
sessions. condition be experienced the same number of
Response definitions, measurement, and IOA. times on a given day. Participants were required
As in Study 1, we recorded all sessions using to experience a minimum of five sessions in each
camcorders or other recording equipment, and condition, and the comparison was considered
data were scored using laptops equipped with complete when this requirement was met and
rates of problem behavior and independent that were synthesized during IISCA-based FCT
FCR use appeared stable in at least one condi- could all be withheld at the same time
tion for at least three consecutive sessions. (e.g., both escape extinction and
tangible extinction may be implemented
simultaneously).4
Procedures Emily. For both FCT conditions, the EO
FCT developed from the IISCA involved involved presenting Emily with the demand
teaching an omnibus FCR that produced 30 s to brush her hair or unpack her backpack.
of access to the multiple reinforcers for problem When this evocative event was presented
behavior identified in the IISCA (Hanley et al., during IISCA-based FCT sessions, saying
2014). FCT developed from the standard FA “my way please” produced 30 s of escape
involved teaching an FCR that produced 30 s from these demands, access to preferred
of access to the isolated reinforcer for problem music (tangible), and access to child-directed
behavior identified in the standard FA. We sig- interaction (attention). After 30 s, the music
naled the different FCT conditions with color- was paused, the analyst stopped engaging in
correlated cards approximately 22 x 28 cm (for Emily-led interactions, and Emily was
participants who communicated vocally) or instructed to return to work. When the
color-correlated iPad™ buttons approximately same EO was presented during standard FA-
1.5 x 1.5 cm (for participants who communi- based FCT sessions, saying “take a break
cated with an SGD). All sessions began with please” produced 30 s of escape from
the analyst showing the participant the corre- demands. No music was provided, the task
lated stimulus and describing the contingency materials were removed, and the analyst did
in place for that session. not provide Emily with any attention. In
Each trial consisted of the following compo- addition, the analyst stepped back from
nents: (a) the analyst implemented the evocative Emily approximately .6 m to make the deliv-
event that had produced problem behavior in ery of escape more salient.
the FA upon which the treatment was based, Jeff. Both of Jeff’s FCRs were produced
(b) the participant emitted the corresponding using his SGD, which was an iPad™ equipped
FCR, and (c) the analyst reinforced the FCR by with the LAMP Words for Life™ communica-
delivering the reinforcer(s) that had been deliv- tion app and encased in a Big Grips® foam case
ered in that particular FA. For participants who for protection. Both FCRs required Jeff to push
communicated vocally, a model prompt was ini- individual buttons for each word, and both
tially provided to occasion the FCR. For partici- FCRs required an equal number of button
pants who communicated with an SGD, a presses. Gesture prompts were provided one at
gesture prompt to push the correct buttons in a time (i.e., the analyst gestured to “my” and
the correct order was initially provided. The waited for Jeff to push that button, then ges-
delay to prompts was faded within session: 0 s, tured to “way” and waited for Jeff to push that
3 s, 5 s. Each instance of the FCR produced button). These gesture prompts were faded on
30 s of reinforcement whether it occurred inde-
pendently or with a prompt. The prompt delay 4
In Emily’s case in which attention, tangible, and
was increased each time the participant com- escape functions were combined, it may seem that escape
pleted 4–5 trials at the current delay, until extinction and attention extinction cannot be implemen-
reaching a terminal delay of 5 s. Problem behav- ted simultaneously. However, the specific type of atten-
tion being evaluated for Emily was child-directed
ior and precursors in both conditions were interaction, which was easily withheld while presenting
placed on extinction; the particular reinforcers demands.
a delay as described above. An independent, he had to get his temperature taken (with no
correct response was scored if Jeff pushed the toys present). The analyst approached Dylan
buttons in the correct order and then pushed with a thermometer to place on his forehead.
the top of the screen to voice the entire phrase, Saying “not now” produced 30 s of escape
with no more than 5 s pause between button from the demand to get his temperature
presses and without any prompts from the taken.
analyst. Chloe. The EO for both conditions involved
The EO for IISCA-based FCT involved the analyst attempting to help Chloe put her
pausing Jeff’s YouTube video and instructing shoes on. During IISCA-based FCT, saying
him to return to his vocational counting task. “my way please” produced 30 s of escape from
Saying “my way” produced 30 s of escape from the analyst’s prompts to put the shoes on, as
the vocational task plus access to videos on well as access to engaging in ritualistic behav-
YouTube. The EO for standard FA-based FCT ior with the shoes. During standard FA-based
involved the analyst telling Jeff that she had FCT, saying “take a break please” produced
some work to do, and then turning away from 30 s of escape only. The analyst moved
him. Saying “let’s talk” produced 30 s of atten- Chloe’s shoes to the side and did not provide
tion from the analyst in the form of talking access to engaging in ritualistic behavior with
about any topics Jeff initiated. Jeff communi- the shoes. Attention was not provided during
cated primarily with his SGD, but could pro- this escape interval, and all requests to engage
duce vocal approximations of many words. He in ritualistic behavior with the shoes were
initiated topics by producing a word approxi- ignored.
mation, such as “pi-ta” for “pizza.” The analyst
would then talk about pizza with Jeff, and Jeff
responded by interjecting “yeah” or “okay” as Results
the analyst spoke, or answering simple ques- IISCA-based FCT was effective for all four
tions (e.g., saying “mom” when the analyst participants (Figures 4 and 5). It was more
asked who Jeff was going to eat pizza with effective than standard FA-based FCT for
later). Emily and Jeff; it was equally effective as stand-
Dylan. Dylan experienced three FCT condi- ard FA-based FCT for Chloe and Dylan. For
tions because his standard FA indicated control Emily, problem behavior was eliminated during
by two separate reinforcers. We will refer to IISCA-based FCT after one session. The FCR
these latter two conditions as tangible-based my way please began occurring independently
FCT and escape-based FCT throughout the during the second session, and was acquired by
rest of the article. The EO for IISCA-based the third session. For Jeff, no problem behavior
FCT involved telling Dylan he needed to put occurred during any IISCA-based FCT session;
away his toys and come get his temperature the FCR began occurring independently during
taken. Saying “my way” produced 30 s of the first session and was acquired by the fifth
escape from the thermometer and access to session. By contrast, problem behavior persisted
the preferred toys. The EO for tangible-based during standard FA-based FCT for both of
FCT involved telling Dylan he had to surren- these participants, and neither participant
der his toys. The analyst approached him with acquired the standard FA-based FCR. It should
an outstretched hand to take his toys (with no be noted that both Emily and Jeff did engage
thermometer present). Saying “toys please” in many prompted instances of the standard
produced 30 s of access to the toys. The EO FA-based FCR, and thus they contacted the
for escape-based FCT involved telling Dylan reinforcement contingency many times each
BL BL
5 IISCA Standard FCT + EXT 2 IISCA Standard FCT + EXT
Combined PB and
4 Escape to music, Escape to YouTube

PC per min
child-directed interaction Attention

3 Escape
1
2
1
0 0
2
per min
FCR
Emily Jeff
0
5 10 15 5 10 15 20
Sessions
Figure 4. Rate of combined problem behavior and precursors and independent FCRs during IISCA-based FCT and
standard FA-based FCT treatment for Emily (left panels) and Jeff (right panels).
session. However, independent use of the DISCUSSION

standard FA-based FCR was never established. The results of our Study 1 revealed that when
In both conditions, any FCRs emitted were rel- implemented as they have been most often
evant to the condition (i.e., the IISCA-based described in the literature, the IISCA was differ-
FCR never occurred during standard FA-based
entiated more often than the standard FA (100%
sessions, and vice versa).
versus 44% of applications) and was more effi-
Both treatments were equally effective in
cient. Our results regarding the speed and effi-
eliminating problem behavior and establishing
cacy of the IISCA in demonstrating a functional
FCRs for Dylan and Chloe (see Figure 5). In
Dylan’s case, no problem behavior occurred relation between problem behavior and environ-
during IISCA-based FCT or tangible-based mental events are similar to those reported by
FCT; minimal problem behavior occurred others who have evaluated the IISCA
during one session of escape-based FCT. In (Ghaemmaghami et al., 2015, 2016; Hanley
Chloe’s case, problem behavior (grabbing the et al., 2014; Jessel et al., 2016; Santiago et al.,
analyst) initially occurred during both FCT 2016). We compared function-based treatments
conditions, but was reduced to zero by the in Study 2 and found that IISCA-based treat-
eighth session in each condition. All FCRs ment was effective in all four cases and standard
were easily acquired by the second session for FA-based treatment was similarly effective in two
both participants and occurred at similar rates cases. Taken together, the results from Study
across conditions. In addition, both Dylan and 1 and Study 2 suggest that the IISCA can be an
Chloe exclusively used the FCR relevant to effective format for assessing and treating socially
each condition. mediated problem behavior.
BL
BL
rd
5 IISCA Standard FCT + EXT FCT + EXT
Combined PB and
A
nda
C
IIS
Sta
4
PC per min
Escape to tangibles Escape to rituals

3 Escape Escape
Tangible
2
1
0
2 5
4
per min
FCR
3
1
2
Dylan 1 Chloe
0 0
5 10 15 20 5 10 15 20 25 30
Sessions
Figure 5. Rates of combined problem behavior and precursors and independent FCRs during IISCA-based FCT
and standard FA-based FCT treatment for Dylan (left panels) and Chloe (right panels).
Regarding Study 1, there are several possible may be that problem behavior for some partici-
reasons the IISCA was differentiated more pants was maintained by interactions between
often than the standard FA. The first and most contingencies. For example, Diego’s outcomes
obvious may be that precursors were reinforced show that his problem behavior was evoked by
in the IISCA and not in the standard a synthesis of EOs and maintained by a synthe-
FA. There were five participants for whom the sis of reinforcers, rather than by any one EO or
standard FA was undifferentiated; we examined consequence in isolation. The fact that problem
the possibility for each of them. For Diego and behavior for Jonah was controlled and differen-
Jonah, precursors did not occur or were not tiated in the IISCA but not in the standard FA
reported. Thus, the lack of differentiation in also suggests the potential presence of impor-
their standard FAs was not related to the exclu- tant interactions. Jonah’s IISCA contained an
sion of precursors. For Mason and Riley, the arrangement of variables capable of reliably
exclusion of precursors was directly responsible evoking problem behavior as well as reliably
for the lack of differentiation in their standard suppressing it; by contrast, the arrangement of
FAs. We conducted precursor FAs for both of variables in the standard FA did not produce
these participants, allowing us to evaluate control. There are a number of other reported
response rates when the same contingencies of instances in which a synthesis of contingencies
the standard FA were applied to precursors reliably controlled problem behavior when iso-
observed in the IISCA, and saw differentiated lated contingencies failed to do so
rates when precursors were reinforced in the (e.g., Adelinis & Hagopian, 1999; Bowman
standard FA. et al., 1997; Fisher et al., 1998; Ghaemma-
A second possible reason that the IISCA was ghami et al., 2015; Hanley et al., 2014;
differentiated more often than the standard FA Mann & Mueller, 2009; Mueller et al., 2005;
Sarno et al., 2011). The results for Diego and two analyses to determine the extent to which
Jonah in Study 1 highlight the importance of the IISCA detected the same reinforcers as the
considering interactions in addition to main standard FA. In other words, these authors
effects when analyzing problem behavior. One used the standard FA as an answer key of sorts
advantage afforded by the IISCA may be its against which to evaluate the validity of the
ability to evoke problem behavior that is sensi- IISCA. This interpretation assumes that the
tive only to interactions. standard FA is a truth-bearing criterion that
The presence of interactions between contin- provides the correct answer regarding the rein-
gencies may have also contributed to the fact forcers that maintain problem behavior, and
that IISCA-based FCT was more effective than analyses that provide different information
standard FA-based FCT for Emily and Jeff in must therefore be incorrect. However, rather
Study 2. It is possible that interactions were than assume the veracity of one analysis based
present and went undetected even though their on the fact that it has been applied more often,
standard FAs were differentiated (for other our results from Study 2 suggest that a more
examples, see Mann & Mueller, 2009; Payne pragmatic strategy is to compare the relative
et al., 2014). For example, Emily’s standard FA efficacy of treatments designed from sets of dif-
indicated a clear escape function, and no prob- ferentiated analyses. The IISCA-based treat-
lem behavior or precursors occurred in any ment eliminated problem behavior in all cases
other condition. However, despite these clear and established an appropriate alternative
results, a treatment based on escape as the response in all cases. These factors are especially
maintaining consequence did not eliminate important given that a primary purpose of an
problem behavior. The addition of music and FA is to develop an effective function-based
child-directed interaction during the reinforce- treatment for problem behavior. We think this
ment interval was necessary for treatment pragmatic approach is preferable because it
effects to be observed with Emily judges the value of an analysis by the extent to
(as demonstrated in the IISCA-based treat- which its corresponding treatment produces
ment). The fact that no problem behavior meaningful outcomes. We caution against dis-
occurred in the attention or tangible condition missing the utility of the IISCA based on the
of her standard FA indicates that these EOs conclusion that its results do not match those
were not sufficient to evoke problem behavior of the standard FA, particularly given that the
in and of themselves, but it does not rule out data from Study 2 indicate that the IISCA pro-
the possibility that these events interact with duced effective treatment more often than the
other reinforcers to maintain problem behavior. standard FA.
Emily’s treatment data make it clear that music The tactic of examining treatment data to
and child-directed interaction were in fact rele- affirm (or reject) FA conclusions has been suc-
vant to her problem behavior, despite the fact cessfully used by several other authors. For
that they were not detected by the instance, Smith, Iwata, Vollmer, and Zarcone
standard FA. (1993) noted that problem behavior occurring
Our results from Study 1 are similar to those in multiple test conditions could reflect mainte-
of Fisher et al. (2016) in that both analyses nance by multiple reinforcers or a failure to iso-
detected similar though not identical reinforcers late the necessary variables. These authors
for most participants. However, our interpreta- sought to affirm the functions detected in the
tion of those results differs significantly. Fisher analysis by implementing treatments based on
et al. provided an interpretation of IISCA and each function and found that problem behavior
standard FA outcomes based on comparing the was in fact multiply maintained for two of
three participants. Payne et al., (2014) reported contexts in which it occurred, that the assess-
two cases in which treatment based on the ment produced a useful baseline against which
results of an FA was not effective in eliminating to evaluate treatment effects, and that “support
problem behavior. Subsequent analyses deter- for the synthesized contingency analyses….
mined that some interactions between reinfor- comes from the meaningful outcomes produced
cers had not been detected by the FA for all three participants” (p. 33). In other
(e.g., problem behavior was maintained by words, the primary reason these authors offered
escape to attention), and treatment modifica- for synthesizing contingencies is that emulating
tions based on these combined reinforcers were ecologically relevant scenarios can assist in pro-
found to be effective. In a third example, ducing large, generalizable, and socially vali-
Ghaemmaghami et al. (2015) conducted an dated effects.
IISCA that synthesized the reinforcers of Hanley et al. (2014) do point out the impor-
escape, attention, and tangible items, and tance of the synthesized contingency for partici-
found that problem behavior was eliminated pant Gail in that her problem behavior
during treatment when and only when mands occurred when and only when contingencies
for all reinforcers were acquired, affirming that were synthesized, and they suggest that reinfor-
the reinforcers included in the IISCA were all cing effects of contingencies can be missed
relevant parts of the contingency maintaining when evaluated in isolation. The independent
problem behavior. effects of each reinforcer were not evaluated for
There are several additional interpretations participants Bob and Dale in Hanley et al., but
offered by Fisher et al. (2016) that are impor- their treatments were highly effective and
tant to discuss: (a) maintenance by interactive socially validated anyway. We found that a syn-
effects is the underlying assumption of the thesis of reinforcers was not necessary to pro-
IISCA; (b) the IISCA is therefore validated duce effective treatment for two of four
only when response patterns presumably indic- participants (Dylan and Chloe), yet their IIS-
ative of interactive effects are observed between CAs were differentiated and were completed in
the IISCA and standard FA; (c) including extra fewer sessions than the standard
stimuli that do not function as reinforcers in FA. Additionally, IISCA-based treatment was
the IISCA can be problematic because it may effective even with the inclusion of possibly
lead to an unnecessarily cumbersome treat- irrelevant contingencies, which more closely
ment. There are a number of conceptual and mimicked the situations reported by caregivers
practical problems with these premises and the as problematic. Thus, an important point
conclusions that follow from them. regarding the utility of the IISCA is that it has
The IISCA does not assume maintenance by been shown to produce effective treatment
interactive effects. Hanley et al. (2014) did not regardless of whether problem behavior is main-
describe this as the rationale for the assessment tained by interactive effects (Gail in Hanley
and treatment model, nor did other authors et al.; Emily in the current study), independent
who have reported using IISCAs effects (Dylan, Chloe in the current study), or
(e.g., Ghaemmaghami et al., 2015, 2016; Jessel whether the independent effects are unknown
et al., 2016; Santiago et al., 2016). The (Bob, Dale in Hanley et al.). More impor-
description of the IISCA as assuming interac- tantly, our results for Emily and Jeff suggest
tive effects is unique to Fisher et al. (2016). In that the standard FA can sometimes miss
their discussion, Hanley et al. note that a syn- potential interactions and therefore produce
thesis of contingencies allowed problem behav- treatments that are not effective, even when the
ior to be quickly evaluated in the typical FA is differentiated. This outcome seems more
problematic than a treatment including poten- study is that continued attempts to modify and
tially incidental stimuli that do not function as refine the standard FA may have produced dif-
reinforcers. ferentiation and effective treatment for more
Regarding the concern that including inci- participants (as was evident in Hagopian et al.).
dental stimuli in an IISCA may produce a more However, it is important to consider that the
cumbersome treatment, this did not appear to process of modifying and re-running the analy-
be the case for Dylan or Chloe in Study sis is time consuming and may require a level
2. Dylan’s escape-based FCT treatment was of expertise that is not easily accessible outside
actually more cumbersome to implement of specialized hospital or clinic settings. As an
because we had to clear the session room of all example, the BCBAs who worked with our par-
toys to ensure that he could not wander around ticipants as part of their school program had an
the room and access preferred toys during his average caseload of 10 students to oversee,
escape interval. Typical school, clinic, and which amounts to about 3 hours per student
home environments are not devoid of toys or per week given a 30-hour school week. It is
other tangible items, and it is not likely that possible that conducting three iterations of a
parents or practitioners would specifically pro- standard FA could take 3 or more weeks when
hibit access to toys during treatment based on operating under those time constraints. Instead
an escape function. Thus, whether or not tangi- of conducting multiple iterations of an analysis
ble items are relevant to the maintenance of to obtain differentiation, an alternative course
problem behavior, it may actually be less cum- could be to design individualized test-control
bersome and more practical to provide escape analyses from open-ended interviews at the
to tangibles (instead of escape to nothing) dur- start (Hanley, 2012; Hanley et al., 2014). Our
ing treatments for escape-maintained problem results suggest that synthesizing multiple vari-
behavior. In addition, although Dylan’s ables suspected of influencing problem behavior
tangible-based FCT condition was effective in is likely to produce an efficient analysis that is
eliminating problem behavior, it did not differentiated on the first attempt (Study 1)
address the need for caregivers to be able to and a treatment that is also effective (Study 2).
take his temperature (a primary concern of his Although we found the IISCA to be differ-
caregivers). A second treatment would be neces- entiated and to produce effective treatment
sary to establish compliance with this and other more often than the standard FA, we do not
related demands, which would necessitate that mean to suggest that it can or should now sup-
caregivers be trained in two different treat- plant the standard FA under all conditions.
ments. This seems more cumbersome than a The development of the standard FA was the
synthesized treatment shown to be equally catalyst for a critical paradigm shift to function-
effective. based treatment for problem behavior; as a
It is interesting to note that the percentage research platform, it allows behavior analysts to
of differentiation we obtained with the standard continue evaluating isolated components of
FA (44%) is similar to that reported by Hago- analyses and treatments to determine their
pian et al. (2013) in which only 47% of stand- value. In their discussion, Hanley et al. (2014)
ard FAs were differentiated on the first attempt. emphasized the bidirectional nature of studies
There are a considerable number of ways in that synthesize multiple assessment or treat-
which the standard FA may be modified when ment components and studies that evaluate
differentiated results are not initially obtained these variables in isolation. In particular, these
(such as including precursors in the contin- authors noted that empirical evaluations of
gency class), and one limitation of the present synthesized treatments can suggest additional
independent variables to be more closely exam- schedules of reinforcement. Our conclusions

ined in isolation, and these types of highly ana- regarding the acceptability of including inciden-
lytic examinations can in turn yield isolated tal stimuli in synthesized treatments must
variables that comprise the constituent parts of therefore be tempered by the short duration
synthesized treatment packages with long-term, over which treatment was applied. Extending
generalized, meaningful outcomes. The present the comparison beyond the very first stages of
studies occasion a number of questions along intervention is a critical next step.
both such lines of research. Finally, the continued evaluation of the gen-
For instance, it may be important to isolate erality of the IISCA will be important, as there
the effects of synthesized response topographies are currently only a handful of published stud-
from the effects of synthesized consequences. ies demonstrating its use, and even fewer stud-
These effects could be separated from each ies comparing it and its derived treatments to
other by conducting IISCAs and standard FAs other FA formats and their derived treatments.
in which identical topographies are reinforced In particular, studies focusing on the long-term
in each analysis (as we did for five of our nine efficacy, generality, and social acceptability of
participants). It may also be important to iso- treatments relying on isolated versus synthe-
late the effects of synthesized versus isolated sized contingencies will be most important in
EOs. If multiple EOs are presented in a test determining the relative benefits and costs of
condition but only one reinforcer is provided focusing on these different types of precision
(e.g., escape), it is difficult to determine when analyzing and treating problem behavior.
whether problem behavior was evoked by the Considering that many practitioners report
EO for that particular reinforcer, by the EO for rarely conducting FAs (Ellingson, Miltenber-
another reinforcer, or by some combination of ger, & Long, 1999; Oliver, Pratt, & Normand,
EOs. Future research could address this ques- 2015; Roscoe, Phillips, Kelly, Farber, & Dube,
tion by comparing conditions in which multi- 2015) and given that research on the IISCA is
ple EOs are arranged and a single reinforcer is only beginning, it will be important to con-
provided (e.g., Call & Lomas Mevers, 2014; tinue to evaluate the extent to which the IISCA
Call, Wacker, Ringdahl, & Boelter, 2005), or any other FA format is efficient, effective,
multiple EOs are arranged and all reinforcers and readily adopted by practitioners in the
are provided (e.g., an IISCA), or an isolated coming years.
EO is arranged by providing continuous access
to all reinforcers but one (e.g., Gail’s isolated
analyses in Hanley et al., 2014). REFERENCES
An important area of research related to the
Adelinis, J. D., & Hagopian, L. P. (1999). The use of
continued evaluation of synthesized analyses symmetrical “do” and “don’t” requests to interrupt
and treatments is to determine the impact of ongoing activities. Journal of Applied Behavior
including incidental stimuli in synthesized con- Analysis, 32, 519–523. doi:10.1901/jaba.1999.
32-519
tingencies during treatment (i.e., including sti-
Beavers, G. A., Iwata, B. A., & Lerman, D. C. (2013).
muli whose independent effects were not Thirty years of research on the functional analysis of
demonstrated). Will doing so impede treatment problem behavior. Journal of Applied Behavior
or enhance it? The inclusion of incidental sti- Analysis, 46, 1–21. doi:10.1002/jaba.30
muli did not seem problematic for Dylan and Bloom, S. E., Iwata, B. A., Fritz, J. N., Roscoe, E. M., &
Carreau, A. B. (2011). Classroom application of a
Chloe’s treatments. However, we did not evalu-
trial-based functional analysis. Journal of Applied
ate their treatments over longer periods of time, Behavior Analysis, 44, 19–32. doi:10.1901/jaba.2011.
in additional contexts, or under thinner 44-19
Borrero, C. S. W., & Borrero, J. C. (2008). Descriptive analysis and treatment of destructive behavior main-
and experimental analyses of potential precursors to tained by termination of “don’t” (and symmetrical
problem behavior. Journal of Applied Behavior “do”) requests. Journal of Applied Behavior Analysis,
Analysis, 41, 83–96. doi:10.1901/jaba.2008.41-83 31, 339–356. doi:10.1901/jaba.1998.31-339
Bowman, L. G., Fisher, W. W., Thompson, R. H., & Fisher, W. W., Greer, B. D., Romani, P. W.,
Piazza, C. C. (1997). On the relation of mands and Zangrillo, A. N., & Owen, T. M. (2016). Compari-
the function of destructive behavior. Journal of sons of synthesized and individual reinforcement con-
Applied Behavior Analysis, 30, 251–265. doi:10.1901/ tingencies during functional analysis. Journal of
jaba.1997.30-251 Applied Behavior Analysis, 49, 596–616. doi:10.1002/
Call, N. A., & Lomas Mevers, J. E. (2014). The relative jaba.314
influence of motivating operations for positive and Fisher, W. W., Piazza, C. C., Cataldo, M. F., Harrell, R.,
negative reinforcement on problem behavior during Jefferson, G., & Conner, R. (1993). Functional com-
demands. Behavioral Interventions, 29, 4–20. munication training with and without extinction and
doi:10.1002/bin.1374 punishment. Journal of Applied Behavior Analysis, 26,
Call, N. A., Wacker, D. P., Ringdahl, J. E., & 23–36. doi:10.1901/jaba.1993.26-23
Boelter, E. W. (2005). Combined antecedent vari- Fisher, W. W., Piazza, C. C., & Chiang, C. L. (1996).
ables as motivating operations within functional ana- Effects of equal and unequal reinforcer duration dur-
lyses. Journal of Applied Behavior Analysis, 38, ing functional analysis. Journal of Applied Behavior
385–389. doi:10.1901/jaba.2005.51-04 Analysis, 29, 117–120. doi:10.1901/jaba.1996.
Call, N. A., Wacker, D. P., Ringdahl, J. E., Cooper- 29-117
Brown, L., & Boelter, E. W. (2004). An assessment Fritz, J. N., Iwata, B. A., Hammond, J. L., &
of antecedent events influencing noncompliance in Bloom, S. E. (2013). Experimental analysis of precur-
an outpatient clinic. Journal of Applied Behavior sors to severe problem behavior. Journal of Applied
Analysis, 37, 145–157. doi:10.1901/jaba.2004. Behavior Analysis, 46, 101–129. doi:10.1002/jaba.27
37-145 Ghaemmaghami, M., Hanley, G. P., & Jessel, J. (2016).
Camp, E. M., Iwata, B. A., Hammond, J. L., & Contingencies promote delay tolerance. Journal of
Bloom, S. E. (2009). Antecedent versus consequent Applied Behavior Analysis, 49, 548–75. doi:10.1002/
events as predictors of problem behavior. Journal of jaba.33
Applied Behavior Analysis, 42, 469–483. doi:10.1901/ Ghaemmaghami, M., Hanley, G. P., Jin, S. C., &
jaba.2009.42-469 Vanselow, N. R. (2015). Affirming control by multi-
Campbell, J. M. (2003). Efficacy of behavioral interven- ple reinforcers via progressive treatment analysis.
tions for reducing problem behavior in persons with Behavioral Interventions, 31, 70–86. doi:10.1002/
autism: A quantitative synthesis of single-subject bin.1425
research. Research in Developmental Disabilities, 24, Hagopian, L. P., Rooker, G. W., Jessel, J., &
120–138. doi:10.1016/S0891-4222(03)00014-3 DeLeon, I. G. (2013). Initial functional analysis out-
Carr, E. G., & Durand, V. M. (1985). Reducing behavior comes and modifications in pursuit of differentiation:
problems through functional communication train- A summary of 176 inpatient cases. Journal of Applied
ing. Journal of Applied Behavior Analysis, 18, Behavior Analysis, 46, 88–100.
111–126. doi:10.1901/jaba.1985.18-111 Hall, S. S. (2005). Comparing descriptive, experimental
Derby, K. M., Wacker, D. P., Sasso, G., Steege, M., and informant-based assessments of problem beha-
Northup, J., Cigrand, K., & Asmus, J. (1992). Brief viors. Research in Developmental Disabilities, 26,
functional assessment techniques to evaluate aberrant 514–526. doi:10.1016/j.ridd.2004.11.004
behavior in an outpatient setting: A summary of Hanley, G. P. (2012). Functional assessment of problem
79 cases. Journal of Applied Behavior Analysis, 25, behavior: Dispelling myths, overcoming implementa-
713–721. doi:10.1901/jaba.1992.25-713 tion obstacles, and developing new lore. Behavior
Ellingson, S. A., Miltenberger, R. G., & Long, E. S. Analysis in Practice, 5, 54–72.
(1999). A survey of the use of functional assessment Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003).
procedures in agencies serving individuals with devel- Functional analysis of problem behavior: A review.
opmental disabilities. Behavioral Interventions, 14, Journal of Applied Behavior Analysis, 36, 147–185.
187–198. doi:10.1002/(SICI)1099-078X(199910/ doi:10.1901/jaba.2003.36-147
12)14:4<187::AID-BIN38>3.0.CO;2-A Hanley, G. P., Jin, C. S., Vanselow, N. R., &
Falcomata, T. S., Roane, H. S., Feeney, B. J., & Hanratty, L. A. (2014). Producing meaningful
Stephenson, K. M. (2010). Assessment and treatment improvements in problem behavior of children with
of elopement maintained by access to stereotypy. autism via synthesized analyses and treatments. Jour-
Journal of Applied Behavior Analysis, 43, 513–517. nal of Applied Behavior Analysis, 47, 16–36. doi:10.10
doi:10.1901/jaba.2010.43-513 02/jaba.106
Fisher, W. W., Adelinis, J. D., Thompson, R. H., Hawkins, R. P. (1979). The functions of assessment:
Worsdell, A. S., & Zarcone, J. R. (1998). Functional Implications for selection and development of devices
for assessing repertoires in clinical, educational, and with autism. Journal of Applied Behavior Analysis, 42,
other settings. Journal of Applied Behavior Analysis, 355–360. doi:10.1901/jaba.2009.42-355
12, 501–516. doi:10.1901/jaba.1979.12-501 Lalli, J. S., Mace, F. C., Wohn, T., & Livezey, K. (1995).
Herscovitch, B., Roscoe, E. M., Libby, M. E., Identification and modification of a response-class
Bourret, J. C., & Ahearn, W. H. (2009). A proce- hierarchy. Journal of Applied Behavior Analysis, 28,
dure for identifying precursors to problem behavior. 551–559. doi:10.1901/jaba.1995.28-551
Journal of Applied Behavior Analysis, 42, 697–702. Langdon, N. A., Carr, E. G., & Owen-DeSchryver, J.
doi:10.1901/jaba.2009.42-697 (2008). Functional analysis of precursors for serious
Higgins Hains, A., & Baer, D. M. (1989). Interaction problem behavior and related intervention. Behavior
effects in multielement designs: Inevitable, desirable, Modification, 32, 804–827. doi:10.1177/01454455
and ignorable. Journal of Applied Behavior Analysis, 08317943
22, 57–69. doi:10.1901/jaba.1989.22-57 Leon, Y., Lazarchick, W. N., Rooker, G. W., &
Iwata, B. A., DeLeon, I. G., & Roscoe, E. M. (2013). DeLeon, I. G. (2013). Assessment of problem behav-
Reliability and validity of the functional analysis ior evoked by disruption of ritualistic toy arrange-
screening tool. Journal of Applied Behavior Analysis, ments in a child with autism. Journal of Applied
46, 271–284. doi:10.1002/jaba.31 Behavior Analysis, 46, 507–511. doi:10.1002/jaba.41
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Lerman, D. C., & Iwata, B. A. (1993). Descriptive and
Bauman, K. E., & Richman, G. S. (1994). Toward a experimental analyses of variables maintaining self-
functional analysis of self-injury. Journal of Applied injurious behavior. Journal of Applied Behavior
Behavior Analysis, 27, 197–209. doi:10.1901/jaba. Analysis, 26, 293–319. doi:10.1901/jaba.1993.
1994.27-197 26-293
Iwata, B. A., & Dozier, C. L. (2008). Clinical application Mace, F. C., & Lalli, J. S. (1991). Linking descriptive
of functional analysis methodology. Behavior Analysis and experimental analyses in the treatment of bizarre
in Practice, 1, 3–9. speech. Journal of Applied Behavior Analysis, 24,
553–562. doi:10.1901/jaba.1991.24-553
Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R.,
Magee, S. K., & Ellis, J. (2000). Extinction effects during
Vollmer, T. R., Smith, R. G., … Willis, K. D.
the assessment of multiple problem behaviors. Journal
(1994). The functions of self-injurious behavior: An
of Applied Behavior Analysis, 33, 313–316. doi:10.19
experimental-epidemiological analysis. Journal of
01/jaba.2000.33-313
Applied Behavior Analysis, 27, 215–240. doi:10.1901/
Mann, A. J., & Mueller, M. M. (2009). False positive
jaba.1994.27-215
functional analysis results as a contributor of treat-
Iwata, B. A., Wong, S. E., Riordan, M. M., ment failure during functional communication train-
Dorsey, M. F., & Lau, M. M. (1982). Assessment ing. Education & Treatment of Children, 32,
and training of clinical interviewing skills: Analogue 121–149. doi:10.1353/etc.0.0044
analysis and field replication. Journal of Applied Moore, J. W., Mueller, M. M., Dubard, M.,
Behavior Analysis, 15, 191–203. doi:10.1901/jaba. Roberts, D. S., & Sterling-Turner, H. (2002). The
1982.15-191 influence of therapist attention on self-injury during
Jessel, J., Hanley, G. P., & Ghaemmaghami, M. (2016). a tangible condition. Journal of Applied Behavior
Interview-informed synthesized contingency analysis: Analysis, 35, 283–286. doi:10.1901/
Thirty replications and reanalysis. Journal of Applied jaba.2002.35-283
Behavior Analysis, 49, 576–595. doi:10.1002/jaba. Mueller, M. M., Sterling-Turner, H., & Moore, J. W.
316 (2005). Towards developing a classroom-based func-
Kahng, S., Iwata, B. A., & Lewin, A. B. (2002). Behav- tional analysis condition to assess escape-to-attention
ioral treatment of self-injury, 1964 to 2000. American as a variable maintaining problem behavior. School
Journal on Mental Retardation, 107, 212–221. Psychology Review, 34, 425–431.
doi:10.1352/0895-8017(2002)107<0212:BTOSIT>2. Northup, J., Wacker, D., Sasso, G., Steege, M.,
0.CO;2 Cigrand, K., Cook, J., & DeRaad, A. (1991). A brief
Kodak, T., Northup, J., & Kelley, M. E. (2007). An eval- functional analysis of aggressive and alternative behav-
uation of the types of attention that maintain prob- ior in an outclinic setting. Journal of Applied Behavior
lem behavior. Journal of Applied Behavior Analysis, Analysis, 24, 509–522. doi:10.1901/jaba.1991.
40, 167–171. doi:10.1901/jaba.2007.43-06 24-509
Kuhn, D. E., Hardesty, S. L., & Luczynski, K. (2009). Oliver, A. C., Pratt, L. A., & Normand, M. P. (2015). A
Further evaluation of antecedent social events during survey of functional behavior assessment methods
functional analysis. Journal of Applied Behavior used by behavior analysts in practice. Journal of
Analysis, 42, 349–353. doi:0.1901/jaba.2009.42-349 Applied Behavior Analysis, 48, 817–829. doi:10.1002/
Kuhn, D. E., Hardesty, S. L., & Sweeney, N. M. (2009). jaba.256
Assessment and treatment of excessive straightening Payne, S. W., Dozier, C. L., Neidert, P. L.,
and destructive behavior in an adolescent diagnosed Jowett, E. S., & Newquist, M. H. (2014). Using
additional analyses to clarify the function of problem Smith, R. G., Iwata, B. A., Vollmer, T. R., &
behavior: An analysis of two cases. Education & Zarcone, J. R. (1993). Experimental analysis and
Treatment of Children, 37, 249–275. doi:10.1353/ treatment of multiply controlled self-injury. Journal of
etc.2014.0017 Applied Behavior Analysis, 26, 183–196. doi:10.1901/
Pelios, L., Morren, J., Tesch, D., & Axelrod, S. (1999). jaba.1993.26-183
The impact of functional analysis methodology on St. Peter, C. C., Vollmer, T. R., Bourret, J. C.,
treatment choice for self-injurious and aggressive Borrero, C. S. W., Sloman, K. N., & Rapp, J. T.
behavior. Journal of Applied Behavior Analysis, 32, (2005). On the role of attention in naturally occur-
185–195. doi:10.1901/jaba.1999.32-185 ring matching relations. Journal of Applied Behavior
Richman, D. M., Wacker, D. P., Asmus, J. M., Analysis, 38, 429–443. doi:10.1901/jaba.2005.
Casey, S. D., & Andelman, M. (1999). Further anal- 172-04
ysis of problem behavior in response class hierarchies. Strohmeier, C. W., Murphy, A., & O’Connor, J. T.
Journal of Applied Behavior Analysis, 32, 269–283. (2016). Parent-informed test-control functional anal-
doi:10.1901/jaba.1999.32-269 ysis and treatment of problem behavior related to
Rispoli, M., Camargo, S., Machalicek, W., Lang, R., & combined establishing operations. Developmental
Sigafoos, J. (2014). Functional communication train- Neurorehabilitation. Advance online publication.
ing in the treatment of problem behavior maintained doi:10.3109/17518423.2015.1133723
by access to rituals. Journal of Applied Behavior
Analysis, 47, 580–593. doi:10.1002/jaba.130 Thomason-Sassi, J., Iwata, B. A., Neidert, P. L., &
Roscoe, E. M., Kindle, A. E., & Pence, S. T. (2010). Roscoe, E. M. (2011). Response latency as an index
Functional analysis and treatment of aggression main- of response strength during functional analyses of
tained by preferred conversational topics. Journal of problem behavior. Journal of Applied Behavior
Applied Behavior Analysis, 43, 723–727. doi:10.1901/ Analysis, 44, 51–67. doi:10.1901/jaba.2011.44-51
jaba.2010.43-723 Thompson, R. H., & Iwata, B. A. (2007). A comparison
Roscoe, E. M., Phillips, K. M., Kelly, M. A., of outcomes from descriptive and functional analyses
Farber, R., & Dube, W. V. (2015). A statewide sur- of problem behavior. Journal of Applied Behavior
vey assessing practitioners’ use and perceived utility Analysis, 40, 333–338. doi:10.1901/jaba.2007.56-06
of functional assessment. Journal of Applied Behavior Tiger, J. H., Fisher, W. W., Toussaint, K. A., &
Analysis, 48, 830–844. doi:10.1002/jaba.259 Kodak, T. (2009). Progressing from initially ambigu-
Santiago, J. L., Hanley, G. P., Moore, K., & Jin, C. S. ous functional analyses: Three case examples. Research
(2016). The generality of interview-informed func- in Developmental Disabilities, 30, 910–926. doi:10.10
tional analyses: Systematic replications in school and 16/j.ridd.2009.01.005
home. Journal of Autism and Developmental Disorders, Tiger, J. H., Hanley, G. P., & Bruzek, J. (2008). Func-
46, 797–811. doi:10.1007/s10803-015-2617-0 tional communication training: A review and practi-
Sarno, J. M., Sterling, H. E., Mueller, M. M., cal guide. Behavior Analysis in Practice, 1, 16–23.
Dufrene, B., Tingstrom, D. H., & Olmi, D. J.
Wallace, M. D., & Iwata, B. A. (1999). Effects of session
(2011). Escape-to-attention as a potential variable for
duration on functional analysis outcomes. Journal of
maintaining problem behavior in the school setting.
Applied Behavior Analysis, 32, 175–183. doi:10.1901/
School Psychology Review, 40, 57–71.
jaba.1999.32-175
Schlichenmeyer, K. J., Roscoe, E. M., Rooker, G. W.,
Wheeler, E. E., & Dube, W. V. (2013). Idiosyncratic Zarcone, J. R., Fisher, W. W., & Piazza, C. C. (1996).
variables that affect functional analysis outcomes: A Analysis of free time contingencies as positive versus
review (2001–2010). Journal of Applied Behavior negative reinforcement. Journal of Applied Behavior
Analysis, 46, 339–348. Analysis, 29, 247–250. doi:10.1901/jaba.1996.
Smith, R. G., & Churchhill, R. M. (2002). Identification 29-247
of environmental determinants of behavior disorders
through functional analysis of precursor behaviors. Received February 7, 2016
Journal of Applied Behavior Analysis, 35, 125–136. Final acceptance October 17, 2016
doi:10.1901/jaba.2002.35-125 Action Editor, James Carr

Slaton Et Al-2017-Journal of Applied Behavior Analysis

Uploaded by

Copyright:

Available Formats

Slaton Et Al-2017-Journal of Applied Behavior Analysis

Uploaded by

Document Information

Original Title

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

Slaton Et Al-2017-Journal of Applied Behavior Analysis

Uploaded by

Copyright:

Available Formats

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2017, 50, 252–277 NUMBER 2 (SPRING)

INTERVIEW-INFORMED FUNCTIONAL ANALYSES: A COMPARISON

Participant Problem Behaviors (reported to occur together) Precursor(s)

IISCA Standard FA Third Analysis

4 Escape to music, Escape to YouTube

child-directed interaction Attention

session. However, independent use of the DISCUSSION

Escape to tangibles Escape to rituals

independent variables to be more closely exam- schedules of reinforcement. Our conclusions

You might also like