You are on page 1of 22

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2012, 45, 1–22 NUMBER 1 (SPRING 2012)

FUNCTIONAL ANALYSIS AND TREATMENT OF ARRANGING AND


ORDERING BY INDIVIDUALS WITH AN AUTISM
SPECTRUM DISORDER
NICOLE M. RODRIGUEZ AND RACHEL H. THOMPSON
WESTERN NEW ENGLAND COLLEGE

KEVIN SCHLICHENMEYER
NEW ENGLAND CENTER FOR CHILDREN

AND

COREY S. STOCCO
WESTERN NEW ENGLAND COLLEGE

Of the diagnostic features of autism, relatively little research has been devoted to restricted and
repetitive behavior, particularly topographically complex forms of restricted and repetitive
behavior such as rigidity in routines or compulsive-like behavior (e.g., arranging objects in
patterns or rows). Like vocal or motor stereotypy, topographically complex forms of restricted
and repetitive behavior may be associated with negative outcomes such as interference with skill
acquisition, negative social consequences, and severe problem behavior associated with
interruption of restricted and repetitive behavior. In the present study, we extended functional
analysis methodology to the assessment and treatment of arranging and ordering for 3
individuals with an autism spectrum disorder. For all 3 participants, arranging and ordering was
found to be maintained by automatic reinforcement, and treatments based on function reduced
arranging and ordering.
Key words: arranging and ordering, autism, cleaning, completeness, compulsive behavior,
washing, functional analysis, higher level restricted and repetitive behavior
________________________________________

Compared with the other diagnostic features sive-like1 behavior such as arranging objects in
of autism, little research has been devoted to patterns or rows) are understudied (Bodfish,
restricted and repetitive behavior (Bodfish, 2004; Turner, 1999). This lack of research is
2004; South, Ozonoff, & McMahon, 2005; problematic because, like vocal or motor
Turner, 1999). In particular, topographically stereotypy, more complex forms of restricted
complex forms of restricted and repetitive and repetitive behavior may be associated with
behavior (e.g., rigidity in routines or compul- negative outcomes such as interference with skill
acquisition (e.g., Dunlap, Dyer, & Koegel,

This study was conducted in partial fulfillment of the


1
first author’s requirements for the doctoral degree at The terms compulsive-like behavior or compulsive
Western New England College. We thank Gregory P. behavior will be used to refer to responses that share
Hanley, Amanda M. Karsten, and William H. Ahearn for topographical similarities with the behavior of individuals
their thoughtful comments on an earlier version of this with obsessive compulsive disorder and for the purpose of
manuscript. maintaining consistency with the terminology used in the
Correspondence concerning this article should be RBS-R scale (Bodfish, Symons, & Lewis, 1999). In
addressed to Nicole M. Rodriguez, who is now at the particular, compulsive-like behavior will be used when
Munroe-Meyer Institute, 985450 Nebraska Medical referring to any of the responses included in the
Center, Omaha, Nebraska 68198 (e-mail: nicole. compulsive behavior subscale of the RBS-R scale (e.g.,
rodriguez@unmc.edu). arranging and ordering, completeness, or washing and
doi: 10.1901/jaba.2012.45-1 cleaning).

1
2 NICOLE M. RODRIGUEZ et al.

1983), negative social consequences (Jones, restricted and repetitive behavior, we could
Wint, & Ellis, 1990), and severe problem not locate any studies in the autism literature
behavior associated with its interruption (Cuc- that sought to treat problematic forms of higher
caro et al., 2003; Flannery & Horner, 1994; level restricted and repetitive behavior.
Turner, 1999). In fact, several researchers have One productive approach to addressing prob-
demonstrated a functional relation between ag- lem behavior is to develop treatments based on
gression and disruption and access to the oppor- its function. Iwata, Dorsey, Slifer, Bauman, and
tunity to engage in compulsive-like behavior Richman (1982/1994) developed a comprehen-
with individuals with autism (e.g., Hausman, sive functional analysis to test behavioral sensi-
Kahng, Farrell, & Mongeon, 2009; Kuhn, tivity to three main types of reinforcement
Hardesty, & Sweeney, 2009) and other devel- contingencies: social positive, social negative,
opmental disabilities (e.g., Murphy, Macdon- and automatic. Originally applied to the assess-
ald, Hall, & Oliver, 2000). ment and treatment of self-injurious behavior,
As a means of distinguishing between forms functional analyses since have been applied to the
of restricted and repetitive behavior that vary in study of a wide range of problem behavior,
topographical complexity, Turner (1999) sug- including aggression (e.g., Fisher et al., 1993),
gested subdividing restricted and repetitive pica (e.g., Piazza et al., 1998), property destruc-
behavior into higher level and lower level classes. tion (e.g., Fisher, Lindauer, Alterson, & Thomp-
According to Turner, higher level restricted and son, 1998), and lower level restricted and
repetitive behavior includes complex behavior repetitive behavior, such as motor stereotypy
such as circumscribed interests (e.g., preoccu- (e.g., Kennedy, Meyer, Knowles, & Shukla,
pation with serial numbers on electronics), rigid 2000). In addition to demonstrating the gener-
and invariant routines (e.g., dressing, eating, or ality of functional analysis methodology, these
playing in a particular pattern), and arranging studies demonstrated how functional analyses
and ordering (e.g., lining things up in patterns can be adjusted to allow the assessment of
or rows); lower level restricted and repetitive behavior other than self-injury.
behavior includes less complex forms of stereo- To date, no published studies have used
typy such as repetitive motions (e.g., hand functional analysis methodology to analyze higher
flapping) and repetitive manipulation of objects level restricted and repetitive behavior. Thus, the
(e.g., object spinning). Several hypotheses have purpose of the present study was to extend the
been proposed as to the function of higher level functional analysis model to the assessment and
restricted and repetitive behavior. For example, treatment of arranging and ordering and other
this type of behavior has been said to be compulsive-like behavior (e.g., completeness2 and
evidence of a ‘‘need for sameness’’ (e.g., Kanner, washing and cleaning) for three individuals with
1943; Prior & MacMillan, 1973) or an inherent an autism spectrum disorder. Arranging and
lack of behavioral flexibility (Green et al., ordering was selected as the topography of interest
2006), yet few studies have sought to under- based on results from an unpublished survey that
stand environmental variables that contribute to we conducted (data available from the first
this apparent need for sameness or lack of author), which identified arranging and ordering
behavioral flexibility (cf. Flannery & Horner, as one of the most problematic forms of higher
1994, who assessed the role of predictability on level repetitive behavior for the population
problem behavior associated with disruption of
2
routines). Further, even though individuals with See Response Measurement and Interobserver Agree-
ment section for operational definition of completeness.
an autism spectrum disorder are commonly The terms completeness or complete items will be used to
described as engaging in both classes of refer to responses outlined in this operational definition.
ANALYSIS AND TREATMENT OF ARRANGING 3

sampled (14% of 102 students with an autism threads; see below for operational definitions).
spectrum disorder reportedly engaged in arrang- Teachers reported that Ross excessively straight-
ing and ordering that was rated as severely ened or organized items (e.g., on shelves, in
problematic). Participants in this study represent closets, or in refrigerators) or fixed anything
those students for whom the frequency and that was out of place (e.g., a loose thread on
severity of their arranging and ordering and other someone’s shirt or a book placed upside down
related compulsive-like behavior warranted inter- on a bookshelf). His arranging and ordering
vention. often caused delays when transitioning to the
next scheduled activity. For example, when
GENERAL METHOD retrieving a drink from the refrigerator during
dinner, Ross often arranged items in the
Participants refrigerator until someone interrupted him.
All three participants attended a specialized Among the more problematic forms of Ross’s
school and residential program for individuals arranging and ordering was his arranging of
with autism and other developmental disabilities. items in trash bins, which posed health risks.
Jim was a 15-year-old boy who had been His behavior also extended to other people (e.g.,
diagnosed with pervasive developmental disor- fixing buttons or loose threads on others’
der (not otherwise specified). He followed clothing), which teachers reported could cause
simple instructions and spoke in one- to problems if it involved another student. Picking
three-word sentences but with a limited vocab- of loose threads on his own clothing sometimes
ulary. His compulsive-like behavior included resulted in large holes in his clothing. In
arranging and ordering (e.g., aligning objects or addition, interruption of compulsive-like behav-
furniture against other objects or surfaces) and ior sometimes resulted in minor self-injurious
completeness (e.g., insisting that items such as behavior, property destruction, tantrums, or
drawers or doors remained completely closed; vocal perseveration on the item until it was
see below for operational definitions). Jim also fixed.
engaged in high rates of lower level restricted Christie was a 15-year-old girl who had been
and repetitive behavior including vocal stereo- diagnosed with autism. She followed simple
typy and repetitive touching or tapping of items instructions and spoke using two- to four-word
and surfaces. Collectively, his restricted and sentences. Her compulsive-like behavior includ-
repetitive behavior occurred at such a high rate ed arranging and ordering of furniture and
that his teachers described him as constantly other items (e.g., magazines or leisure items).
moving or engaging in some form of restricted Teachers reported that interrupting or blocking
and repetitive behavior. Teachers reported that arranging and ordering of furniture or other
his behavior was difficult to interrupt or items often resulted in aggression or self-
redirect, that the severity of this behavior made injurious behavior. Among the most problem-
it difficult for them to complete academic work atic forms of arranging and ordering was her
with him, and that he would sometimes push arranging of large pieces of furniture, which
his teacher out of the way to arrange and order could be dangerous should, for example, the
items. furniture become unstable and fall.
Ross was a 13-year-old boy who had been
diagnosed with autism. He followed instruc- Response Measurement and
tions and spoke in one- to three-word sentences. Interobserver Agreement
His compulsive-like behavior included arrang- Data were collected on arranging and
ing and ordering, completeness, and washing ordering for all participants. When applicable,
and cleaning (e.g., picking at lint or loose data also were collected on other types of higher
4 NICOLE M. RODRIGUEZ et al.

level restricted and repetitive behavior listed on hand. Completeness and washing and cleaning
the compulsive subscale of the Repetitive were recorded using a frequency measure.
Behavior Scale–Revised (RBS-R; developed by When applicable during treatment evalua-
Bodfish, Symons, Parker, & Lewis, 2000; also tions, data also were collected on item engage-
see Bodfish, Symons, & Lewis, 1999), including ment, prompts, response blocking or interrup-
completeness (Jim and Ross) and washing and tion, and product replacement. Duration of
cleaning (Ross only). Response definitions from item engagement was defined differently for
the RBS-R were modified based on information each item but generally included item contact
gathered from interviews and observations to plus appropriate manipulation (not including
encompass each individual’s specific form of the stereotypic manipulation of items). Response
target behavior. blocking or interruption (hereafter referred to as
Arranging and ordering was defined as moving response blocking) was scored when the ex-
objects from one location to another (when not perimenter placed her hand between the par-
related to retrieving or putting away leisure or ticipant and the materials involved in the target
academic materials or interacting with items as behavior. To accommodate responding during
the manufacturer intended); aligning objects response-blocking conditions, response defini-
against another surface such as a wall, corner, tions for each target behavior included blocked
or another object; and lining up or stacking attempts to engage in that response.
objects when doing so was not the intended Duration data were summarized as percent-
purpose for those objects (e.g., dominos or age duration per session by dividing the number
blocks). Arranging and ordering was recorded of seconds in which the target response occurred
using a duration measure for Ross and Christie. by the total number of seconds in a session
However, because Jim repeatedly aligned the (600 s) and then multiplying the quotient by
same object in the same manner in rapid 100%. Frequency data were summarized as
succession (e.g., placing the trash bin in the responses per minute by dividing the frequency
corner of the room four times in a row within of the target response in a session by the total
4 s), his arranging and ordering was measured number of minutes in a session (10 min).
according to the frequency of episodes. Specif- Sessions were videotaped so that they could
ically, a new episode was scored after a 1-s pause be scored at a later time. Data during all anal-
from the last episode or when he made contact yses and treatment evaluations were collected on
with a new item. For Christie, the functional handheld computers. A second observer inde-
analysis included both arranging and ordering of pendently scored data during a minimum of
furniture and other items such as books and toys. 33% of sessions, distributed equally across
However, because of the difference in the severity conditions. Interobserver agreement for dura-
of the problem across topographies, arranging tion and frequency measures was determined by
and ordering of furniture was recorded separately partitioning sessions into 10-s bins and com-
from other items for all subsequent analyses. paring data collectors’ observations on an
Completeness was defined as closing drawers, interval-by-interval basis. For each interval, the
closing books, and placing tops on containers smaller number (duration or frequency) was
(not including instances during which it would divided by the larger number. The quotient
be natural to do so, such as after retrieving an then was multiplied by 100% and averaged
item from a drawer or reading a book). Washing across all intervals. Intervals during which both
and cleaning was defined as picking at lint or data collectors agreed that the target behavior
loose threads, picking up trash from the floor or did not occur were factored into the above
other surfaces, and wiping surfaces with one’s equation as 100% agreement for that interval.
ANALYSIS AND TREATMENT OF ARRANGING 5

Mean agreement scores for Jim’s experimental conducted in session rooms (1.5 m by 3 m).
analyses were 93% (range, 67% to 100%) for Each of the session rooms contained a desk, two
arranging and ordering, 97% (range, 89% to chairs, a trash bin, a set of drawers, several
100%) for completeness, 82% (range, 68% to notebooks and magazines, two stacks of mis-
92%) for item engagement, 94% (range, 87% matched cups, a silverware tray with unorga-
to 100%) for prompts, and 97% (range, 93% to nized plastic silverware, academic materials
100%) for response blocking. Mean agreement (e.g., money, colored cards), and at least one
scores for Ross’s experimental analyses were container of items. On the desk was a pencil
90% (range, 67% to 100%) for arranging and holder with pencils and several other items such
ordering, 99.5% (range, 95% to 100%) for as a ruler, cups, or a water bottle. Items in the
completeness, 97% (range, 72% to 100%) for session room were intentionally unaligned (e.g.,
washing and cleaning, 98% (range, 85% to the trash bin was near the wall but placed at an
100%) for item engagement, and 99% (range, angle and the desk was against the wall but with
97% to 100%) for response blocking. Mean approximately 3 cm of space between the desk
agreement scores for Christie’s experimental and wall), mismatched (e.g., several plastic
analyses were 94% (range, 76% to 100%) for blocks were placed in a bin of wooden blocks),
arranging and ordering, 95% (range, 83% to placed outside their bins, or were incomplete
100%) for item engagement, and 92% (range, (e.g., magazines and notebooks were left open
77% to 100%) for prompts. and drawers and the top of the container were
left slightly ajar). This was done to allow mul-
FUNCTIONAL ANALYSES tiple opportunities to arrange and order or com-
plete items. For Ross, we added enough items
Given the paucity of studies that have (e.g., crumpled paper, cups, napkins) to the
addressed arranging and ordering in autism, trash bin to fill it to the rim, three loose threads
we began by gathering information regarding placed on various parts of the experimenter’s
the topography of interest via teacher interviews clothing, and a plant (Ross arranged and
and naturalistic observations (data available ordered the dirt, stems, and leaves of plants).
from the first author). This information was Sessions for Christie were conducted in the
used to arrange experimental analyses such that living room of her residence. The room con-
the types of antecedent and consequent events
tained a locked TV cabinet, two couches, an
programmed matched those typically encoun-
end table, a wooden desk and chair, two book-
tered during naturally occurring conditions
shelves, two locked cabinets, a treadmill, an area
(thereby increasing the ecological validity of
carpet (1.8 m by 2.5 m), a table (0.6 m by
our experimental analyses; Hanley, Iwata, &
1.8 m) with various items (e.g., magazines and
McCord, 2003). Operational definitions and
dolls) on it, a basket of toys, a large bouncy ball,
session materials were also informed by inter-
views and observations in natural settings. All and four magazines on the floor. Videos were
sessions lasted 10 min. played on the TV during all sessions to approx-
imate conditions under which arranging and
Comprehensive Functional Analysis ordering occurred during naturalistic obser-
Assessments for each participant began with a vations. No other students were present during
functional analysis of arranging and ordering sessions.
and other related compulsive-like behavior. The experimenter developed the initial
Settings and materials. The setting in which arrangement of furniture and materials for each
the functional analysis was conducted varied for participant to permit numerous opportunities
each participant. Sessions for Jim and Ross were to observe the target behavior. Although the
6 NICOLE M. RODRIGUEZ et al.

number of items that could be completed or An escape condition was included to identify
washed and cleaned were limited compared to whether the targeted restricted and repetitive
those items that could be arranged and ordered, behavior functioned to postpone or terminate a
materials (aside from leisure items in the nonpreferred activity. Demands were selected
attention and control condition) were held based on interviews and descriptive data and
constant across sessions to ensure equal oppor- were delivered using three-step guided compli-
tunity to engage in target behavior across ance (i.e., vocal, gesture or model, and physical
sessions. It also should be noted that partici- prompts, with 5 s between each prompt).
pants were never observed to produce an en- Contingent on target behavior, the experiment-
tirely orderly environment prior to the end of a er turned away from the participant and
session, which suggests that the opportunity to provided a 30-s break from the task. Demand
engage in target behavior was present through- materials were removed only when they were
out each session. not involved in target behavior. For Jim and
Procedure. Functional analysis procedures Ross, an avoidance contingency also was in-
were similar to those described by Iwata et al. cluded, in which the escape interval was ex-
(1982/1994); however, antecedent and conse- tended by 30 s contingent on a target behavior
quent events were adjusted to accommodate (e.g., Iwata et al. 1982/1994). The inclusion of
information gathered from interviews and nat- an avoidance contingency was consistent with
uralistic observations. Three test conditions naturalistic observations, during which Jim and
(attention, escape, and no interaction) and a Ross sometimes engaged in compulsive-like
control condition were alternated in a multi- behavior prior to transitioning to a new location
element design. Different-colored discrimina- and activity, which may have delayed the onset
tive stimuli (i.e., poster board, t-shirt, or of the next activity. In such cases, stimuli other
tablecloth) were paired with each condition of than an explicit instruction may have signaled
the functional analysis to facilitate discrimina- the onset of a new activity (e.g., when a teacher
tion between contingencies that operated in said to another teacher, ‘‘We’re on our way to
different conditions (Conners et al., 2000). the academic classroom.’’).
In the attention condition, two low-prefer- Although interviews and descriptive data for
ence leisure items were available while the Christie did not suggest that arranging and
experimenter pretended to be occupied reading ordering occurred in the context of demands,
a magazine (leisure items were identified using we included an escape condition because escape
paired-item preference assessment procedures is the most common maintaining variable for
described by Fisher et al., 1992; data available other problem behavior of individuals with
from the first author). Contingent on target developmental disabilities and related disorders
behavior, the experimenter provided attention (e.g., Iwata, Pace, Dorsey, et al., 1994). To
on a fixed-ratio (FR) 1 schedule. Based on ensure that there was an opportunity to arrange
narrative records from naturalistic observations, furniture, we included demands that allowed
attention for Jim and Ross took the form of Christie to move about the room such as
comments about the ordering (e.g., ‘‘It’s fine physical education (e.g., jumping jacks) and
Jim, you don’t need to fix it.’’) or reprimands clean-up tasks (e.g., wiping windows). An
(e.g., ‘‘Stop!’’ ‘‘Leave it.’’). Attention for avoidance contingency was not included in the
Christie took the form of vocal reprimands escape condition for Christie because, unlike for
(e.g., ‘‘Christie, stop!’’) and statements of Jim and Ross, naturalistic observations did not
concern (e.g., ‘‘Careful, it could fall on you. suggest that arranging and ordering may have
You could hurt yourself.’’). served to delay the onset of demands (e.g.,
ANALYSIS AND TREATMENT OF ARRANGING 7

arranging and ordering was not observed to pleteness and washing and cleaning occurred in
occur following the delivery of instructions). the no-interaction condition, suggesting that
We included a no-interaction condition, completeness and washing and cleaning likely
during which the experimenter was in the room were maintained by automatic reinforcement.
but did not interact with the participant, to Christie’s levels of arranging and ordering
identify whether the target restricted and repeti- (bottom panel) were variable across all condi-
tive behavior persisted in the absence of social tions, except the escape condition, which was
contingencies. A no-interaction condition was relatively low and stable. In addition, respond-
included instead of an alone condition for safety ing persisted across both series of no-interaction
reasons and because some target responses sessions, suggesting that her arranging and ordering
required the presence of another person (e.g., was maintained by automatic reinforcement.
removing threads from other people’s clothing).
In the control condition, no demands were Follow-Up Functional Analyses
presented and attention was delivered every 30 s Escape analysis (Ross). Ross’s functional anal-
and contingent on requests for attention (e.g., ysis suggested that arranging and ordering was
vocal requests for tickles). However, if a target maintained by both automatic reinforcement
behavior occurred immediately prior to a and escape from demands. However, it is possi-
ble that responding in the escape condition was
scheduled attention delivery, attention was
not due to the avoidance or escape contingency
delayed for 5 s. In addition, two highly preferred
but, instead, to the lack of alternative stimula-
leisure items, selected based on the results of a
tion available during avoidance or escape inter-
paired-item preference assessment (Fisher et al.,
vals (Kuhn, DeLeon, Fisher, & Wilke, 1999). If
1992), were continuously available.
so, lower levels of responding may have been
Experimental conditions were conducted in a
observed in the control and attention conditions
fixed sequence (i.e., no interaction, attention, because highly preferred leisure items in the
control, and escape; e.g., Iwata, Pace, Dorsey, control condition competed with arranging and
et al., 1994). We conducted several consecutive ordering, and vocal comments in the attention
no-interaction sessions at the end of the analysis condition functioned as punishment. Thus, we
to determine whether responding would persist conducted an analysis similar to that conducted
in the absence of social contingencies, providing by Kuhn et al. (1999) in which a function-based
further support for an automatic function (e.g., treatment that was matched to one function but
Vollmer, Marcus, & LeBlanc, 1994). not the other was used to clarify whether ar-
Results and discussion. Functional analysis ranging and ordering was maintained by both
results for all three participants are depicted in automatic reinforcement and escape from de-
Figure 1. Jim’s levels of responding (top panel) mands or by automatic reinforcement alone.
were variable across all conditions and persisted Setting and design. Ross’s escape analysis was
across a series of no-interaction sessions, suggesting conducted using a multiple baseline design across
that his arranging and ordering and completeness settings. The first setting was the same as the
were maintained by automatic reinforcement. functional analysis. The second setting was a
Ross’s levels of arranging and ordering (second classroom equipped with a table (1.5 m by 1.5 m),
and third panels) were higher in the no- six chairs, a computer and desk, three book-
interaction and escape conditions and persisted shelves containing a variety of books and leisure
across a series of repeated no-interaction condi- items, a locked toy cabinet, a set of drawers left
tions, suggesting that his arranging and ordering slightly ajar, a plant, and a trash bin filled with
was maintained by automatic reinforcement and clean items. As in the functional analysis, items
escape from demands. Higher levels of com- were mismatched, placed outside their bins, or
8 NICOLE M. RODRIGUEZ et al.

Figure 1. Functional analysis data for Jim (top), Ross (second and third panels), and Christie (bottom).
ANALYSIS AND TREATMENT OF ARRANGING 9

unaligned. In addition, three loose threads were ment) was responsible for the decrease in
placed on the experimenter’s clothing. responding and the fact that responding did not
No interaction and escape. The no-interaction recover to baseline levels in the reversal in Setting
and escape conditions from Ross’s functional 1 are both limitations of our escape analysis.
analysis served as the baseline for the escape Process versus product analysis (Christie). We
analysis in the first setting. Baseline in the second conducted a process versus product analysis
setting was the same as the no-interaction con- with Christie to identify the nature of the
dition of the function analysis. automatic reinforcer that maintained arranging
Response blocking and blocking plus escape. and ordering of furniture. There appeared to be
Response blocking was added to both no- consistencies in the direction in which furniture
interaction and escape conditions in an attempt was moved during functional analysis sessions.
to eliminate the automatic reinforcer. If target Thus, we hypothesized that arranging furniture
behavior also was maintained by escape from may have been maintained by some property of
demands, it would persist in the escape con- its final placement (i.e., the product of arranging
dition when attempts to engage in target be- and ordering) rather than the opportunity to
havior resulted in escape but not access to the arrange and order the furniture (i.e., the process
automatic reinforcer. During response blocking, of arranging and ordering). If the final placement
the experimenter blocked attempts to engage in of furniture was the reinforcer, then additional
a target behavior or interrupted a target be- treatment options that did not involve physical
havior after initiation of the response. If Ross contact would be available (e.g., maintaining
successfully displaced an item (which rarely furniture in a preferred placement or replacing
occurred), the item was replaced immediately in furniture contingent on arranging and ordering).
its previous position. The response blocking This was important, because teachers hesitated to
plus escape condition was similar to the escape physically block Christie because it reportedly
condition of the functional analysis except that evoked or elicited aggression.
the target behavior was blocked or interrupted; Setting. Sessions were conducted in the same
attempts to engage in the target behavior con- room as Christie’s functional analysis.
tinued to result in a 30-s break from demands Procedure. The process versus product analy-
or a 30-s delay to the onset of the next demand. sis involved a comparison of two antecedent
Results and discussion. Figure 2 shows the conditions: furniture arranged according to the
results of Ross’s escape analysis. When the original arrangement used in all previous sessions
automatic reinforcer was eliminated, levels of (original arrangement condition) or furniture
arranging and ordering decreased across both arranged according to a preferred arrangement
response blocking and blocking plus escape (preferred product condition). No differential
conditions. That is, responding was not main- consequences were provided contingent on
tained when escape was provided as a reinforcer, arranging and ordering during either condition.
suggesting that arranging and ordering likely was Preferred product probes. Prior to beginning
maintained by automatic reinforcement alone. the process versus product analysis, we con-
An alternative explanation for the decrease in ducted five 50-min probes during which no
responding observed in the escape analysis, differential consequences were provided for
however, is that response blocking may have arranging and ordering and the experimenter
functioned as a punisher. Indeed, this alternate honored all of Christie’s requests for help in
hypothesis is supported by the fact that respond- moving the furniture. During these probes,
ing did not recover to baseline levels in Setting Christie consistently arranged furniture such
1 (third phase, top panel). The difficulty in that the carpet was in the corner by the TV
determining what process (extinction or punish- cabinet, and the couches, bookshelf, table, desk,
10 NICOLE M. RODRIGUEZ et al.

Figure 2. The percentage duration of arranging and ordering during Ross’s escape analysis. Data from his functional
analysis were used as the initial phase in Setting 1 (top). The percentage duration of arranging and ordering in Setting 2 is
depicted in the bottom panel.

chair, and bouncy ball were placed tightly arranging and ordering would be similar to
together in a cluster around the TV cabinet. levels in the original arrangement condition.
Original arrangement. This condition served Prior to preferred product sessions, furniture
as the comparison (or control) condition. The was arranged according to information provided
furniture arrangement at the start of session was by the 50-min probes. In addition, Christie
the same as the functional analysis. was given three 15-s opportunities to request
Preferred product placement. The preferred adjustments. This was done to accommodate
product condition served as the test condition. small adjustments (e.g., aligning the corners of
If arranging and ordering was maintained by the couches) as well as to detect large changes in
access to a product, then little to no arranging preference prior to the start of session. Specif-
and ordering should occur when furniture was ically, the experimenter asked Christie, ‘‘Where
prearranged according to Christie’s preferred would you like the furniture?’’ and pointed to
furniture placement. On the other hand, if several pieces of furniture while asking, ‘‘Stay
arranging and ordering was maintained by the here or move?’’ If Christie requested that
process of arranging and ordering, then levels of something be moved, she was taken to another
ANALYSIS AND TREATMENT OF ARRANGING 11

Figure 3. The percentage duration of arranging and ordering of furniture across preferred product placement and
original arrangement conditions for Christie’s process versus product analysis.

room while the experimenter arranged the establishing operation for arranging and order-
furniture according to her request. This was ing furniture had been eliminated and that the
done in the event that watching the furniture product and not the process of arranging and
being moved was reinforcing. If Christie ordering was the reinforcer.
requested large changes to the arrangement of
the furniture (e.g., moving the TV cabinet from
one side of the living room to the other), TREATMENT EVALUATION
sessions were terminated and a 50-min pre-
Results from functional analyses were used to
ferred product probe was conducted. This
develop function-based treatments for each of
occurred once, before Session 15. Sessions were
the three participants. Due to variations in
never conducted after preferred product probes
in case the provision of 50-min access to procedures across participants, procedures are
arranging and ordering served as an abolishing presented individually for each participant. All
operation. sessions were 10 min.
Results and discussion. Figure 3 shows the
results from Christie’s process versus product Settings
analysis. Levels of arranging and ordering were Sessions for Jim and Christie were conducted
consistently higher in the original arrangement in the same room as their functional analyses.
condition. Because this analysis involved ante- Due to the limited availability of settings
cedent manipulations, the nature of the auto- used during previous analyses, Ross’s treatment
matic reinforcer could be inferred based on evaluation was conducted in the recreational
levels of responding that would be expected in room of his residence. The arrangement of
the presence and absence of an establishing materials in the room was similar to previous
operation for access to a final product. The fact analyses except that a plant was not present, and
that Christie did not arrange furniture in the instead of loose threads on the experimenter’s
preferred product condition suggests that the clothing, four small pieces of paper were
12 NICOLE M. RODRIGUEZ et al.

scattered across the floor (picking up trash from similar to the matched items condition, except
the floor was scored as washing and cleaning). that prompts were provided every 15 s unless
Jim already was engaged with a matched item.
Jim Thus, the number of prompts per session varied.
Baseline. The series of no-interaction sessions Specifically, if he was not engaged with an item
at the end of Jim’s functional analysis served as at the end of a 15-s interval, the experimenter
the initial baseline for his treatment evaluation. delivered a vocal and model prompt (e.g., ‘‘Let’s
Matched items. This condition was included put all the blue pegs over here’’ while modeling
to identify whether providing Jim with appro- the action). If Jim did not engage with a
priate materials to arrange and order and matched item within 2 s of the vocal and model
complete would decrease inappropriate forms prompt, the experimenter provided the least
of arranging and ordering and completeness. amount of physical guidance necessary to initiate
Sessions were similar to baseline except that two item engagement. A voice recorder signaled the
matched items were continuously available. A 15-s intervals.
preliminary competing items assessment (e.g., Matched items plus prompts plus response blocking.
Piazza, Adelinis, Hanley, Goh, & Delia, 2000) A response-blocking component was added to
was not conducted because Jim typically banged, disrupt the relation between arranging and
ripped, or threw leisure items during informal ordering and completeness and the resulting
probes. Instead, matched items were selected automatic reinforcer. Attempts to engage in the
based on whether they (a) could be arranged in a target behavior were blocked by placing the
manner that minimized inappropriate play (e.g., experimenter’s hand between the participant’s
by fastening items to a table) and (b) were likely hand and the materials involved in arranging and
to produce automatic reinforcement similar to ordering and completeness. If the experimenter
that produced by arranging and ordering or could not block the response, the target behavior
completeness. Matched items included a learning was interrupted and the item ordered or completed
cube and a toy bin. The learning cube (0.3 m by was immediately replaced in its original position.
0.3 m by 0.3 m) contained rollercoaster beads, a Matched items plus prompts plus response blocking
sliding peg maze, an abacus, and a xylophone. 2 s. Jim engaged in successive attempts to arrange
The toy bin (0.5 m by 0.5 m by 0.4 m) contained and order or to complete materials during match-
a variety of leisure items (e.g., monster truck and ed items plus prompts plus response-blocking
ball) as well as materials that could be arranged sessions. In an effort to interrupt repeated at-
and ordered or completed but on a smaller, more tempts at the target behavior more effectively, the
contained scale (e.g., miniature drawers and response-blocking component was modified in
containers). The learning cube was secured to this condition such that the experimenter guided
the table using bungee cords, and items in the bin Jim’s hands to his lap for 2 s.
were attached by strings so that Jim could not Teacher implementation. To identify whether
throw or destroy them. the effects of the final intervention package
Matched items plus prompts. Because Jim did would be maintained when implemented by
not engage with the leisure items in the matched Jim’s teachers, two of his teachers (Teachers A
items condition, we conducted separate training and B) implemented the final treatment. All
sessions to teach appropriate item engagement other stimulus conditions remained constant.
(data available from the first author). Prompts Prior to each teacher’s first session, the ex-
then were included within the formal treatment perimenter demonstrated the procedures and
evaluation to test their effect on arranging and answered questions. Subsequent sessions were
ordering and completeness. This condition was preceded by a brief review of the procedures and
ANALYSIS AND TREATMENT OF ARRANGING 13

an opportunity for teachers to ask questions. following (a) 10 s of arranging and ordering, (b) a
Immediate corrective feedback was provided 3-s pause between instances of a target behavior
during sessions as necessary. (criterion for turning off the duration key during
data collection for arranging and ordering was 3 s),
Ross or (c) an attempt to arrange and order, complete,
Although Ross’s escape analysis suggested or wash and clean a new item.
that response blocking was sufficient for Teacher implementation. This condition was the
decreasing his target behavior, we conducted a same as Jim’s teacher implementation condition.
separate treatment evaluation to determine (a) Christie
the influence of competing items on target
Christie’s process versus product analysis
behavior and item engagement and (b) whether identified the product of arranging and ordering
the effects of response blocking would persist as the reinforcer. Because her preferred furniture
under less than optimal procedural integrity. placement was neither practical nor stable, an
Baseline. Baseline was similar to the no- antecedent approach in which furniture remain-
interaction condition of the functional analysis. ed according to her preferred product was not a
Matched items. Sessions were similar to viable treatment option. Thus, we conducted a
baseline except that two matched items were separate treatment evaluation to explore other
available continuously. Matched items were options.
selected based on results of a separate competing Baseline. Baseline was identical to the no-
items assessment (see Piazza et al., 2000). Stimuli interaction condition of the functional analysis.
associated with high levels of item engagement Matched items. Sessions were similar to baseline
and low levels of target behavior included a Lite except that matched items were available contin-
Brite, basketball pinball, and maze. uously. In an attempt to provide materials that
Matched items plus intermittent response blocking. could be arranged and ordered but on a smaller,
Because levels of responding did not recover when more contained scale, we began by including
response blocking was removed in the return to a diorama of a living room with miniature
no-interaction and escape conditions in the escape dollhouse furniture. However, because the doll-
analysis (Figure 2, top), we hypothesized that house furniture did not appear to compete with
responding would be maintained at low levels arranging and ordering, we added three matched
even when response blocking was implemented items (Session 12) based on results of a single-
intermittently. If effective, intermittent response stimulus engagement preference assessment (e.g.,
blocking would serve as a practical alternative DeLeon, Iwata, Conners, & Wallace, 1999). A
to blocking every response, particularly because circus set, garden stringing beads, Beauty and the
Ross typically had a 1:2 teacher-to-student ratio. Beast magnet board, and dollhouse furniture
Sessions were identical to the matched items were associated with the highest levels of item
condition except that the experimenter blocked engagement during the assessment.
the first three of every four attempts to engage in a Matched items plus prompts. This condition was
target behavior (e.g., see Lerman & Iwata, 1996b; similar to the matched items condition except
Smith, Russo, & Le, 1999). Because data on that nondirective vocal and model prompts (e.g.,
arranging and ordering were collected via a Piazza, Contrucci, Hanley, & Fisher, 1997) to
duration measure, we developed criteria for engage with matched items were provided every
initiating response blocking during (or following) 30 s. Specifically, the experimenter modeled
the fourth attempt, during which Ross was appropriate item engagement while describing
allowed to arrange and order. The three-of-four her actions (e.g., ‘‘I’m setting up the couches in
sequence of response blocking was reinitiated front of the TV like a movie theater’’ while
14 NICOLE M. RODRIGUEZ et al.

manipulating the dollhouse furniture) or follow- cumulative item engagement, Christie was
ing a predetermined play script (e.g., ‘‘First up we given 120 s of access to a video that contained
have Pablo on the trapeze!’’ while placing the a collection of preferred movie scenes. This
circus character on the trapeze). reinforcer was selected based on teachers’ reports
Matched items plus prompts plus product ex- that Christie typically watched videos during her
tinction. This condition was similar to the free time and repeatedly requested to watch
matched items plus prompts condition except particular movie scenes in succession. An audible
that the automatic reinforcer associated with timer was used to track her cumulative item
arranging and ordering furniture was eliminated engagement.
(i.e., extinction). Based on the results of the Teacher implementation. This condition was
process versus product analysis, extinction the same as Jim’s teacher implementation
involved immediately replacing furniture in its condition except that the final intervention
original arrangement contingent on arranging package was implemented by three of Christie’s
and ordering (hereafter referred to as product teachers (Teachers A, B, and C).
extinction). On one occasion (Session 36),
Results and Discussion
Christie moved the TV cabinet and then
wedged herself between the cabinet and the Jim. Figure 4 (top) shows the results of
Jim’s treatment evaluation. We began with an
wall such that the experimenter was not able to
assessment of the effects of providing access to
replace the cabinet to its original position. In
matched items amenable to appropriate forms
this situation, the experimenter prevented
of arranging and ordering and completeness.
Christie from moving the cabinet any further
High rates of compulsive-like behavior and low
by placing herself on the other side of the
levels of item engagement suggested that the
cabinet and moving the cabinet to its original automatic reinforcement produced by engaging
location after Christie moved from behind it, with matched items did not compete with the
which occurred at the end of session. As a more automatic reinforcement produced by engaging
conservative measure of the effects of product in arranging and ordering and completeness.
extinction, the time during which Christie Next, we conducted separate training sessions to
physically prevented the experimenter from bring his behavior into contact with the putative
replacing the cabinet was scored as arranging automatic reinforcement produced by engaging
and ordering. with matched items by teaching him how to
Matched items plus prompts plus product engage with leisure items appropriately. When
extinction plus reinforcement for engagement. Al- prompts were introduced into the treatment
though arranging and ordering had decreased to evaluation (matched items plus prompts), the
desirable levels, item engagement remained low. level and variability of arranging and ordering
Thus, we conducted separate training sessions to and completeness decreased and item engage-
teach Christie appropriate item engagement ment increased. Because the rate of prompting
(data available from the first author). Procedures during this condition remained high, it seems
used during play skills training were then added unlikely that the effects of prompts can be
to the treatment evaluation to determine their attributed to increased contact with the putative
effect on item engagement and arranging and automatic reinforcement produced by engaging
ordering. Specifically, nondirective vocal and with matched items. Instead, prompts may have
model prompts were provided (a) at the start of served to interrupt compulsive-like behavior,
session, (b) every 30 s without item engagement, perhaps by diverting attention, albeit temporar-
and (c) immediately after termination of the ily, away from the materials that he typically
reinforcement period. Contingent on 120 s of arranged and ordered or completed.
ANALYSIS AND TREATMENT OF ARRANGING 15

Figure 4. The top panel shows combined arranging and ordering and completeness responses per minute (filled
circles, primary y axis) and the percentage duration of item engagement (open circle, secondary y axis) during Jim’s
treatment evaluation. BL 5 baseline, MI 5 matched items, RB 5 response blocking; the asterisk indicates the point at
which separate play skills training sessions were conducted (between Sessions 12 and 13). Letters in the teacher
implementation phase represent sessions conducted by different teachers (Teacher A and Teacher B). The bottom panel
shows the rate of vocal and model (filled circles) and physical (open circles) prompts.

To further decrease the rate of compulsive- the rate of prompting was on a decreasing trend
like behavior, we introduced response blocking, (bottom panel), and the percentage of item
but the rate of arranging and ordering and engagement was on an increasing trend (top
completeness increased. When the length of panel). It is possible that, with repeated ex-
response blocking was extended to 2 s, rates of posure, engaging with matched items may have
compulsive-like behavior decreased below those begun to take on automatic reinforcement
observed in the matched items plus prompts properties. High levels of item engagement in
condition. Extending the duration of response the absence of prompting would provide further
blocking appeared to interrupt his successive evidence of this possibility. Alternatively, it is
attempts to arrange and order or complete an possible that item engagement served to avoid
item. The effects of the final intervention prompts (prompts were delivered every 15 s
package were maintained when the package unless Jim was already engaged with matched
was implemented by Jim’s teachers. items) and that increases in item engagement
In the final sessions of the teacher imple- can be attributed to negative social reinforce-
mentation condition (Sessions 45 through 48), ment. Additional analyses would be necessary to
16 NICOLE M. RODRIGUEZ et al.

Figure 5. The percentage duration of arranging and ordering and item engagement (top panel; filled circles and open
circles, respectively), completeness, and washing and cleaning responses per minute (middle panel), and frequency of the
experimenter blocking target responses (bottom panel) during Ross’s treatment evaluation. BL 5 baseline, MI 5
matched items, INT RB 5 intermittent response blocking. Letters in the teacher implementation phase of the treatment
evaluation represent sessions conducted by different teachers (Teacher A and Teacher B).

evaluate the behavioral processes that were and ordering decreased when matched items
responsible for increased item engagement. first were introduced; however, these effects
Ross. Figure 5 (top) shows the results of were not replicated when matched items were
Ross’s treatment evaluation. Levels of arranging reintroduced following a reversal to baseline,
ANALYSIS AND TREATMENT OF ARRANGING 17

suggesting that the availability of matched items schedule. By contrast, responding would be
alone did not influence target responding. expected to remain low if punishment is in
Further, with the exception of two sessions, effect. This logic, however, presumes that each
Ross did not engage with matched items during instance of behavior contacts automatic rein-
the matched items condition. Despite the fact forcement. If Ross’s arranging and ordering,
that matched items did not influence target completeness, or washing and cleaning were
behavior, they continued to be made available reinforced, for example, by an orderly environ-
during the remainder of the treatment evalua- ment, then the orderly environment may have
tion to provide Ross with an appropriate not been contacted even when responding was
alternate activity during sessions. The second not blocked. For example, Ross often arranged
matched items condition also served as a the dates that were attached to the calendar via
baseline from which to evaluate the effects of Velcro such that they were aligned and in order.
matched items plus intermittent response block- If the reinforcer was to obtain an orderly calen-
ing, during which levels of arranging and dar, then aligning or ordering only several of the
ordering decreased and rates of completeness dates of the calendar may have not been
and washing and cleaning remained low. sufficient to bring his behavior in contact with
Functional control over the effects of intermit- reinforcement. In other words, extinction may
tent response blocking was demonstrated with have been in effect even when response blocking
arranging and ordering, and the effects of this was intermittent. Thus, the processes that were
treatment persisted when implemented by responsible for the effects of response blocking
Ross’s teachers. Perhaps most notable is that remain unknown.
the frequency of response blocking decreased Christie. Figure 6 (top) shows the results of
and was maintained at low levels by the end of Christie’s treatment evaluation. Levels of ar-
the treatment evaluation. Collectively, these ranging and ordering were variable in the initial
results suggest that the effects of response baseline, with the last three data points ranging
blocking will persist even when implemented from 38% to 44%. The addition of matched
with less than optimal procedural integrity. It items (matched items condition) and prompts
also should be noted that, although minor self- (matched items plus prompts) did not influence
injury, property destruction, and tantrums were levels of arranging and ordering. By contrast,
reported to be associated with interruption of levels of arranging and ordering decreased and
compulsive-like behavior, this behavior was not were maintained at low levels in the matched
observed during the assessment or treatment of items plus prompts plus product extinction
Ross’s compulsive-like behavior in this study. condition. The introduction of supplemental
Lerman and Iwata (1996b) suggested that the reinforcement for engaging with matched items
processes (extinction vs. punishment) that are increased levels of item engagement, and the
responsible for decreased responding can be effects of the final intervention package were
determined by manipulating the schedule of maintained when the treatment was implement-
consequences. If extinction is responsible for ed by Christie’s teachers.
decreased responding when response blocking is Overall, Christie’s results suggested that
implemented on a continuous schedule, behav- product extinction was necessary for maintaining
ior maintained by automatic reinforcement low levels of arranging and ordering. Although
would be expected to increase when response prompts and reinforcement (i.e., access to
blocking is implemented intermittently. In preferred video scenes) also were provided during
other words, responding presumably would her final intervention package to maintain item
contact reinforcement on an intermittent engagement, it seems unlikely that matched
18 NICOLE M. RODRIGUEZ et al.

Figure 6. The top panel shows the percentage duration of arranging and ordering of furniture and the percentage
duration of item engagement during Christie’s treatment evaluation. BL 5 baseline, MI 5 matched items, EXT 5
extinction, Sr+ for ENG 5 reinforcement contingent on item engagement; the asterisk indicates the point at which
separate play skills training was conducted (between Sessions 44 and 45). Letters in the teacher implementation phase
represent sessions conducted by different teachers (Teachers A, B, and C).

items had taken on automatically reinforcing tion of the final treatment in the second to last
properties. Although response blocking should phase and when the final treatment was
serve to disrupt the response–reinforcer relation implemented by Christie’s teachers. We did
regardless of whether the automatic reinforcer is not observe any self-injury throughout Chris-
the process or product of arranging and ordering, tie’s treatment evaluation.
the ability to program an alternate form of Several general limitations of the treatment
extinction (furniture replacement) was particularly evaluations should be noted. First, the effect of
desirable because teachers had hesitated to the final treatment was not evaluated across all
implement procedures that required physical relevant settings in the participants’ natural
contact. Procedural integrity for response blocking environment. Jim’s treatment evaluation was
ultimately may have been compromised if conducted in an analogue setting, and although
Christie’s aggression or self-injury functioned as the treatment evaluation was conducted in
a punisher for teachers’ physical contact. settings frequently encountered by Ross (a
Because interruption of arranging and order- recreational room in his residence) and Christie
ing was reported to be associated with aggres- (the living room of her residence where
sion, we collected data on this behavior during arranging and ordering was reported to be most
all experimental analyses (no differential conse- problematic), compulsive-like behavior also was
quences were provided for nontarget behavior). reported to occur in other settings at their
With few exceptions, each change of condition residence (e.g., kitchen) and school. Second, it
in the treatment evaluation was associated with would have been relevant to determine whether
an initial increase in aggression followed by Jim’s final treatment reduced compulsive-like
decreased levels of responding. Aggression was behavior during instructional contexts, given
maintained at zero levels during implementa- that these contexts were reported to be
ANALYSIS AND TREATMENT OF ARRANGING 19

problematic and levels of compulsive-like operation. Because matched items did not
behavior were relatively high during the escape appear to compete with compulsive-like behav-
condition of his functional analysis. Finally, the ior in our study, it seems unlikely that the
final treatment was relatively labor intensive, motivating operation was addressed. Other
particularly in the components for increasing procedures designed to address the motivating
item engagement for Jim and Christie. Further operation, such as providing an opportunity to
research would be necessary to determine engage in compulsive-like behavior following a
whether procedures applied in the final treat- functional communication response, were not
ment would produce durable behavior change appropriate for these participants, given that the
across all relevant contexts in the participants’ compulsive-like behavior itself was undesirable
natural environment and to what extent the (particularly forms that were considered dan-
intensity of behavioral support could be reduced gerous such as arranging furniture or trash).
over time. Future research might explore acceptable meth-
ods of addressing the motivating operation
GENERAL DISCUSSION when automatically reinforced behavior is
undesirable (e.g., by allowing appropriate forms
Research suggests that extinction may be of compulsive-like behavior).
integral to ensuring treatment success (e.g., We did not conduct a process versus product
Iwata, Pace, Cowdery, & Miltenberger, 1994). analysis for Jim and Ross, in part because
In our study, treatment components aimed at consistencies in the arrangement of items could
disrupting the response–reinforcer relation (re- not be identified easily for all topographies, but
sponse blocking or product extinction) were
also because response blocking prevents access to
necessary to obtain desirable treatment outcomes
a final product placement. Further, aggression or
for all participants. Simply providing access to
self-injury was not reported to be associated with
stimuli that were amenable to appropriate forms
physical contact for Jim or Ross, and response
of compulsive-like behavior (matched items) did
blocking already had been shown to be effective
not reduce the target behavior for any of the
for Ross during his escape analysis. One potential
participants. Further, prompting or prompting
plus supplemental reinforcement was necessary advantage of conducting a process versus product
to maintain engagement with matched items for analysis, even when response blocking is likely to
Jim and Christie, suggesting that these items did be effective, is that the caregiver has slightly
not function as reinforcers. more time to respond to the target behavior
Although response blocking or product ex- when implementing product extinction. If the
tinction was sufficient to reduce compulsive-like compulsive-like behavior is maintained by access
behavior to desirable levels, it often is consid- to a product, any potential detrimental effects of
ered best practice to provide an alternate means procedural integrity errors with response block-
of accessing the functional reinforcer, for ing may be mitigated by immediately replacing
example, by programming its availability via the items in their original positions. Future
differential reinforcement of alternative behav- research may explore this possibility; a paramet-
ior, differential reinforcement of other behavior, ric analysis of various durations of delays to the
or noncontingent reinforcement (Lerman & replacement of items also may be useful for
Iwata, 1996a; Van Houten et al., 1988). In identifying the effects of different types of
addition to providing the individual with the procedural integrity errors and increasing the
opportunity to access personally relevant rein- practicality of treatment.
forcers, these procedures may enhance the ef- The procedures used to identify whether the
fects of treatment by addressing the motivating final placement of items functioned as a reinforcer
20 NICOLE M. RODRIGUEZ et al.

for Christie’s arranging and ordering required petitive behavior is maintained most commonly
that the putative preferred product be identified by automatic reinforcement. Additional research
prior to conducting the process versus product is needed to determine the extent to which higher
analysis. Such antecedent conditions may be level restricted and repetitive behavior in autism
difficult to arrange if consistencies in the final is susceptible to social contingencies.
placement of items cannot be identified (e.g., Arranging and ordering, completeness, and
there was no apparent pattern to Ross’s arranging washing and cleaning in children with autism are
of trash in trash bins), or the estab- commonly described as compulsive behavior
lishing operation for the product as a reinforcer (e.g., Bodfish et al., 1999), perhaps because they
is subtle or difficult to control (e.g., Jim might share topographical similarities with the behavior
dart across the room to close a drawer that was of individuals with obsessive compulsive disorder
only slightly ajar). As noted by Ross’s teacher (OCD). However, it is unknown whether the
during an interview, it would be difficult to avoid compulsive-like behavior of individuals with
or prevent compulsive-like behavior altogether; autism also shares functional similarities with
‘‘[teachers] couldn’t possibly control everything. the compulsive behavior of individuals with
… There is always going to be something on the OCD, which is presumed to be maintained by
floor somewhere or something like a book on a automatic negative reinforcement (i.e., avoid-
shelf that is not facing the correct way.’’ An ance of aversive stimuli; Maltby & Tolin,
alternative means of identifying whether the final 2003). It is interesting to note that exposure
placement of items functions as a reinforcer plus response prevention, which is commonly
would be to assess the effects of product used to treat compulsions in OCD, also shares
extinction. In other words, as is suggested by topographical and functional similarities with
the results of Ross’s escape analysis and several the treatment components found to be effective
other studies that have used similar procedures in the present study (e.g., response blocking). If
(e.g., Kuhn et al., 1999; Smith, Iwata, Vollmer, problem behavior is maintained by automatic
& Zarcone, 1993; Thompson, Fisher, Piazza, & negative reinforcement, then planned and ther-
Kuhn, 1998), function-based treatments can be apist-controlled exposure to aversive stimuli
used to clarify functional relations. Should (e.g., graduated exposure) may enhance the
arranging and ordering persist when items are effects of treatment (Abramowitz, 1996). Fu-
replaced immediately in their original position ture research might evaluate whether the
contingent on arranging and ordering, then the conditions under which compulsive-like behav-
process of arranging and ordering rather than the ior occurs are aversive for individuals with
final placement of items is likely to be the autism (e.g., via a concurrent-chains arrange-
variable that maintains this behavior. ment; see Hanley, Piazza, Fisher, Contrucci, &
We have stressed the role of functional anal- Maglieri, 1997). Other potential sources of
yses in developing treatments, but results from automatic reinforcement, such as ‘‘sameness’’
functional analyses also contribute to our under- (e.g., Prior & MacMillan, 1973), might also be
standing of common maintaining variables. In explored.
their review of the literature on assessment and Although several hypotheses as to the
treatment of stereotypy, Rapp and Vollmer function of higher level restricted and repetitive
(2005) noted that the overwhelming majority behavior have been proposed (e.g., see brief
of the literature suggests that most stereotypy is review by Flannery & Horner, 1994), we
maintained by automatic reinforcement. It is believe that the present study represents the
possible that, like lower level restricted and re- first attempt to identify the function of higher
petitive behavior, higher level restricted and re- level restricted and repetitive behavior in autism
ANALYSIS AND TREATMENT OF ARRANGING 21

using experimental analyses. Results from our Fisher, W. W., Piazza, C., Cataldo, M., Harrell, R., Jefferson,
G., & Conner, R. (1993). Functional communication
study suggest that functional analysis method- training with and without extinction and punishment.
ology, which has been used to assess and treat Journal of Applied Behavior Analysis, 26, 23–36.
lower level restricted and repetitive behavior, Flannery, K. B., & Horner, R. H. (1994). The
relationship between predictability and problem
may be useful for assessing and treating higher behavior for students with severe disabilities. Journal
level restricted and repetitive behavior. Arrang- of Behavioral Education, 4, 157–176.
ing and ordering and other compulsive-like Green, V. A., Sigafoos, J., Pituch, K. A., Itchon, J.,
behaviors were amenable to common function- O’Reilly, M. O., & Lancioni, G. E. (2006). Assessing
behavioral flexibility in individuals with developmen-
based treatment components for all three tal disabilities. Focus on Autism and Other Develop-
participants. Experimental analyses similar to mental Disabilities, 21, 230–236.
those used in this study may inform treatment Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003).
Functional analysis of problem behavior: A review.
of problematic forms of arranging and ordering Journal of Applied Behavior Analysis, 36, 147–185.
or other compulsive-like behavior. Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci,
S. A., & Maglieri, K. A. (1997). Evaluation of client
preference for function-based treatment packages.
REFERENCES Journal of Applied Behavior Analysis, 30, 459–473.
Abramowitz, J. S. (1996). Variants of exposure and Hausman, N., Kahng, S., Farrell, E., & Mongeon, C.
response prevention in the treatment of obsessive- (2009). Idiosyncratic functions: Severe problem
compulsive disorder: A meta-analysis. Behavior Ther- behavior maintained by access to ritualistic behaviors.
apy, 27, 583–600. Education and Treatment of Children, 32, 77–87.
Bodfish, J. W. (2004). Treating the core features of Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E.,
autism: Are we there yet? Mental Retardation and & Richman, G. S. (1994). Toward a functional
Developmental Disabilities, 10, 318–326. analysis of self-injury. Journal of Applied Behavior
Bodfish, J. W., Symons, F. W., & Lewis, M. H. (1999). Analysis, 27, 197–209, (Reprinted from Analysis
The Repetitive Behavior Scale. Western Carolina and Intervention in Developmental Disabilities, 2,
Center Research Reports. 3–20, 1982)
Bodfish, J. W., Symons, F. J., Parker, D. E., & Lewis, M. H. Iwata, B. A., Pace, G. M., Cowdery, G. E., &
(2000). Varieties of repetitive behavior in autism. Journal Miltenberger, R. G. (1994). What makes extinction
of Autism and Developmental Disabilities, 30, 237–243. work: An analysis of procedural form and function.
Conners, J., Iwata, B. A., Kahng, S., Hanley, G. P., Worsdell, Journal of Applied Behavior Analysis, 27, 131–144.
A. S., & Thompson, R. H. (2000). Differential Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R.,
responding in the presence and absence of discriminative Vollmer, T. R., Smith, R. G., et al. (1994). The
stimuli during multielement functional analysis. Journal functions of self-injurious behavior: An experimental-
of Applied Behavior Analysis, 33, 299–308. epidemiological analysis. Journal of Applied Behavior
Cuccaro, M. L., Shao, Y., Grubber, J., Slifer, M., Wolpert, Analysis, 27, 215–240.
C. M., Donnelly, S. L., et al. (2003). Factor analysis Jones, R. S. P., Wint, D., & Ellis, N. C. (1990). The
of restricted and repetitive behaviors in autism using social effects of stereotyped behavior. Journal of
the autism diagnostic interview–Revised. Child Psy- Mental Deficiency Research, 34, 261–268.
chiatry and Human Development, 34, 3–17. Kanner, L. (1943). Autistic disturbances of affective
DeLeon, I. G., Iwata, B. A., Conners, J., & Wallace, contact. Nervous Child, 2, 217–250.
M. D. (1999). Examination of ambiguous stimulus Kennedy, C. H., Meyer, K. A., Knowles, T., & Shukla, S.
preferences with duration-based measures. Journal of (2000). Analyzing the multiple functions of stereo-
Applied Behavior Analysis, 32, 111–114. typical behavior for students with autism: Implica-
Dunlap, G., Dyer, K., & Koegel, R. L. (1983). Autistic tions for assessment and treatment. Journal of Applied
self-stimulation and intertrial interval duration. Behavior Analysis, 33, 559–571.
American Journal of Mental Retardation, 88, 194–202. Kuhn, D. E., DeLeon, I. G., Fisher, W. W., & Wilke, A.
Fisher, W., Lindauer, S. E., Alterson, C. J., & Thompson, E. (1999). Clarifying an ambiguous functional
R. H. (1998). Assessment and treatment of destructive analysis with matched and mismatched extinction
behavior maintained by stereotypic object manipula- procedures. Journal of Applied Behavior Analysis, 32,
tion. Journal of Applied Behavior Analysis, 31, 513–527. 99–102.
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, Kuhn, D. E., Hardesty, S. L., & Sweeney, N. M. (2009).
L. P., Owens, J. C., & Slevin, I. (1992). A comparison Assessment and treatment of excessive straightening
of two approaches for identifying reinforcers for and destructive behavior in an adolescent diagnosed
persons with severe and profound disabilities. Journal with autism. Journal of Applied Behavior Analysis, 42,
of Applied Behavior Analysis, 25, 491–498. 355–360.
22 NICOLE M. RODRIGUEZ et al.

Lerman, D. C., & Iwata, B. A. (1996a). Developing a Smith, R. G., Iwata, B. A., Vollmer, T. R., & Zarcone,
technology for the use of operant extinction in clinical J. R. (1993). Experimental analysis and treatment
settings: An examination of basic and applied research. of multiply controlled self-injury. Journal of Applied
Journal of Applied Behavior Analysis, 29, 345–382. Behavior Analysis, 26, 183–196.
Lerman, D. C., & Iwata, B. A. (1996b). A methodology Smith, R. G., Russo, L., & Le, D. D. (1999).
for distinguishing between extinction and punishment Distinguishing between extinction and punishment
effects associated with response blocking. Journal of effects of response blocking: A replication. Journal of
Applied Behavior Analysis, 29, 231–233. Applied Behavior Analysis, 32, 367–370.
Maltby, N., & Tolin, D. F. (2003). Overview of treatments South, M., Ozonoff, S., & McMahon, W. M. (2005).
for obsessive-compulsive disorder and spectrum con- Repetitive behavior profiles in Asperger syndrome and
ditions: Conceptualization, theory, and practice. Brief high-functioning autism. Journal of Autism and
Treatment and Crisis Intervention, 3, 127–144. Developmental Disorders, 35, 145–158.
Murphy, G., Macdonald, S., Hall, S., & Oliver, C. (2000). Thompson, R. H., Fisher, W. W., Piazza, C. C., & Kuhn,
Aggression and the termination of ‘‘rituals’’: A new D. E. (1998). The evaluation and treatment of
variant of the escape function for challenging behavior? aggression maintained by attention and automatic
Research in Developmental Disabilities, 21, 43–59. reinforcement. Journal of Applied Behavior Analysis,
Piazza, C. C., Adelinis, J. D., Hanley, G. P., Goh, H., & 31, 103–116.
Delia, M. D. (2000). An evaluation of the effects of Turner, M. (1999). Repetitive behaviour in autism: A
matched stimuli on behaviors maintained by auto- review of psychological research. Journal of Child
matic reinforcement. Journal of Applied Behavior Psychology and Psychiatry, 40, 839–849.
Analysis, 33, 13–27. Van Houten, R., Axelrod, S., Bailey, J. S., Favell, J. E.,
Piazza, C. C., Contrucci, S. A., Hanley, G. P., & Fisher, Foxx, R. N., Iwata, B. A., et al. (1988). The right to
W. W. (1997). Nondirective prompting and non- effective behavioral treatment. Journal of Applied
contingent reinforcement in the treatment of destruc- Behavior Analysis, 21, 381–384.
tive behavior during hygiene routines. Journal of Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994).
Applied Behavior Analysis, 30, 705–708. Treatment of self-injury and hand mouthing follow-
Piazza, C. C., Fisher, W. W., Hanley, G. P., LeBlanc, ing inconclusive functional analyses. Journal of
L. A., Worsdell, A. S., Lindauer, S. E., et al. (1998). Applied Behavior Analysis, 27, 331–344.
Treatment of pica through multiple analyses of its
reinforcing functions. Journal of Applied Behavior
Analysis, 31, 165–189.
Prior, M., & MacMillan, M. B. (1973). Maintenance of
sameness in children with Kanner’s syndrome. Journal
of Autism and Childhood Schizophrenia, 3, 154–167.
Rapp, J. T., & Vollmer, T. R. (2005). Stereotypy I: A Received October 23, 2010
review of behavioral assessment and treatment. Final acceptance June 13, 2011
Research in Developmental Disabilities, 26, 527–547. Action Editor, Thomas Higbee

You might also like