Content uploaded by Alan Souza Aranha
Author content
All content in this area was uploaded by Alan Souza Aranha on Oct 14, 2020
Content may be subject to copyright.
263
Clinical
Education
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020.
ISSN 1516-3687 (impresso), ISSN 1980-6906 (on-line). doi:10.5935/1980-6906/psicologia.
v22n3p263-286. Sistema de avaliação: às cegas por pares (double blind review).
Universidade Presbiteriana Mackenzie.
Clinical Psychology
A single-case quasi-experimental
design of Functional Analytic
Psychotherapy for substance abuse
Alan S. Aranha1
http://orcid.org/0000-0002-3230-6827
Claudia K. B. Oshiro¹
http://orcid.org/0000-0001-8965-9124
Elliot C. Wallace2
http://orcid.org/0000-0002-2769-4108
How to cite this article: Aranha, A. S., Oshiro, C. K. B., & Wallace, E. C. (2020). A
single-case quasi-experimental design of functional analytic psychotherapy for sub-
stance abuse. Psicologia: Teoria e Prática, 22(3), 263–286. doi:10.5935/1980-6906/psi-
cologia.v22n3p263-286
Submission: 05/11/2019
Acceptance: 05/27/2020
The content of Psicologia: Teoria e Prática is distributed under the terms of the Creative Commons
Attribution License.
1 University of São Paulo (USP), São Paulo, SP, Brazil.
2 University of Washington (UW), Seattle, Washington – DC, United States.
Clinical
Education
Clinical
Psychology
264 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
Abstract
This study evaluated the eectiveness of Functional Analytic Psychotherapy (FAP) on
clinically relevant behaviors (CRBs) in the context of substance abuse/chemical de-
pendence and identifying the therapeutic components of change. The Functional
Analytic Psychotherapy Rating Scale (FAPRS) was used to categorize therapist and
client behaviors, and Timeline Followback was used to record drug abuse. Two par-
ticipants were treated in a single case A/A + B quasi-experimental design, in which
A = analytical-behavioral therapy and A + B = FAP. The results showed that with the
introduction of FAP, specically the therapist’s contingent responding to the partic-
ipant’s progress, there was a decrease in the use of substances after three months
for both participants. The participant who spent more time in psychotherapy expe-
rienced greater benets at follow-up and a larger reduction in drug use. We conclud-
ed that FAP altered CRBs by positively reinforcing progress in-session and was sig-
nicantly correlated with improvement out-of-session.
Keywords: drug (dependency); behavior therapy; experimental design; drug reha-
bilitation; interpersonal relationships.
UM DELINEAMENTO QUASE-EXPERIMENTAL DE CASO
ÚNICO DA PSICOTERAPIA ANALÍTICA FUNCIONAL PARA
ABUSO DE SUBSTÂNCIAS
Resumo
O objetivo da pesquisa foi avaliar o efeito da Psicoterapia Analítica Funcional (FAP)
sobre os comportamentos clinicamente relevantes (CRBs) e abuso de substâncias de
dependentes químicos e identicar os componentes terapêuticos de mudança.
Foram utilizados os instrumentos Functional Analytic Psychotherapy Rating Scale
para categorização de comportamentos do terapeuta e cliente e Timeline Followback
para registrar o abuso de drogas. Dois participantes foram atendidos em um
delineamento quase-experimental de caso único A/A+B, onde A = Terapia Analítico-
Comportamental e A+B = FAP. Os resultados indicaram que a introdução da FAP, em
especial o responder contingente do terapeuta, acompanhou os progressos e houve
melhora no consumo de substâncias após 3 meses para ambos. O participante que
se manteve mais tempo em psicoterapia obteve maior benefício no follow-up e no
consumo de drogas. Concluiu-se que a FAP alterou CRBs reforçando positivamente
progressos em sessão e que isto provavelmente se correlacionou com melhoras
extrassessão.
Palavras-chave: droga (dependência); terapia comportamental; delineamento ex-
perimental; reabilitação da droga; relações interpessoais.
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 265
FAP and substance abuse
UN DISEÑO DE CASO ÚNICO CUASI EXPERIMENTAL
DE PSICOTERAPIA ANALÍTICA FUNCIONAL PARA ABUSO
DE SUSTANCIAS
Resumen
El objetivo fue evaluar el efecto de la Psicoterapia Analítica Funcional (FAP) en los
comportamientos clínicamente relevantes (clinically relevant behaviors – CRBs) e el
abuso de sustancias de dependientes químicos, e identicar los componentes tera-
péuticos de cambio. Se utilizaron Functional Analytic Psychotherapy Rating Scale,
para clasicar los comportamientos del terapeuta y del cliente, y timeline followback,
para registrar el abuso. Dos participantes fueron evaluados a través de un diseño
cuasiexperimental de caso único A/A + B, donde A = terapia analítico-comporta-
mental y A + B = FAP. Los resultados indicaron que la implementación de la respues-
ta contingente del terapeuta llevó a cambios en las CRBs y mejorías en el consumo
de sustancias en los dos participantes después de tres meses de intervención. El
participante que se mantuvo más tiempo en psicoterapia obtuvo mayor benecio
durante el seguimiento. Se concluye que FAP modicó las CRBs a través del reforza-
miento positivo en sesión, lo que se relacionó con las mejoras fuera de sesión.
Palabras clave: drogas (dependencia); terapia conductista; delineamiento experi-
mental; rehabilitación de la drogadicción; relaciones interpersonales.
1. Introduction
Behavior Analysis (BA) can be divided into three interrelated sub-areas. The
rst, Radical Behaviorism (BR), postures the philosophical aspects of this science,
such as materialism, functionalism, and monism. Experimental Analysis of Behavior
(EAB) is considered basic research, in which a researcher manipulates environmental
variables to understand how they inuence behavior. And nally, Applied Behavior
Analysis (ABA) employs the concepts of basic research to intervene in humans
problems. These three sub-areas are interconnected. EAB and ABA interpret data
under the precepts of RB. The data of EAB is used in new RB reections and in the
services provided by ABA. Results from interventions of ABA strengthen or weaken
assumptions of RB and bring more questions to be answered by EAB (Carvalho
Neto, 2002).
Within ABA, behavioral-analytic psychotherapies are used to evaluate and
intervene with verbal clients. Psychotherapies use functional analysis to identify
266 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
how a client’s behavioral decits and excesses may have been constructed and how
environmental variables maintained them over time. Behavioral intervention
consists of manipulating the maintaining environmental variables to weaken the
problem behaviors and, thus, build more eective behavioral repertoires (Tourinho
& Luna, 2010). One such behavioral psychotherapy, based in the philosophy of RB
and the research method and concepts derived from EAB and ABA, is the Functional
Analytic Psychotherapy (FAP) (Kohlenberg & Tsai, 1991).
1.1 Functional analytic psychotherapy (FAP)
Functional Analytical Psychotherapy’s (FAP) goal is to improve the client’s
interpersonal relationships within their natural environment, which, behaviorally
speaking, can be translated into the development of the interpersonal repertoire.
The goal is to produce positive reinforcers and eliminate aversive social behavior.
To achieve this, FAP utilizes stimulus generalization, stimulus control, and
reinforcement processes. Proponents of this psychotherapy technique posit that
problematic interpersonal response classes, which produce aversive stimulation
and limit access to positive reinforcers, are generalized to the therapeutic
relationship. These behaviors are exhibited in therapy since the psychotherapist
is another person in the client’s life. When problem behaviors occur in a therapy
session, the professional has an opportunity to identify the maintaining
variables, present antecedents that evoke behaviors of interest and present
dierential consequences of problem behaviors. Behavioral improvements are
then expected to re-generalize into an out-of-session context (Kohlenberg &
Tsai, 1991).
FAP classies three types of clinically relevant behaviors (CRB) that can
occur in a psychotherapy session: CRB1 refers to client’s problem behaviors
occurring; CRB2 refers to client’s behaviors that are indicative of therapeutic
progress; and CRB3 are the client’s verbal descriptions of their behavior and the
controlling variables. The therapist’s goal is to decrease the frequency of CRB1s and
increase CRB2s and CRB3s – the latter help the generalization of in-session
changes to out-of-session environments. Kohlenberg and Tsai (1991) describe ve
therapeutic guidelines for a FAP therapist to follow to achieve the goals of the
psychotherapy: rule 1 establishes that the therapist should pay attention to client’s
CRBs during the session; rule 2 involves identifying antecedent events that evoke
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 267
FAP and substance abuse
the occurrence of CRBs; rule 3, the proposed main mechanism of change, states
that the therapist must respond contingently to the client’s CRBs, such as reducing
the frequency of CRB1s (punishment, extinction, blocking avoidance) and increasing
CRB2s and CRBs3 (positive reinforcement); rule 4 involves the therapist evaluating
the impact of their behavior on the client’s repertoire; and rule 5 gives strategies
for generalization outside of therapy.
Initially, FAP was developed to address intimacy and serious interpersonal
problems, such as those found in Axis II of the Diagnostic and Statistical Manual of
Mental Disorders (APA, 2013; Kohlenberg & Tsai, 1991). Subsequently, these
strategies began to be discussed and replicated in diverse types of psychiatric
disorders, theoretical articles, case studies, and experimental research (Kanter et
al., 2017). Two main types of experimental research in particular should be
highlighted: group designs and single-subject designs.
1.2 Experimental research and FAP
In general, experimental research uses scientic methods to assist researchers
in determining the eect of independent variable on the dependent variable,
neutralizing or mitigating the eect of uncontrolled components that can interfere
in the relationship between the independent and dependent variables. Most
experiments have at least two conditions in which the dependent variable results
are compared. By maintaining stability across all variables and manipulating the
independent variable in the experimental condition, it is possible to compare its
eect in relation to the control condition. If there is a measurable dierence,
researchers can, then, conclude that the manipulation of the independent variable
was responsible for the change (Sidman, 1960). We highlight two important
experimental designs: the group designs and single-subject designs.
In group designs, the eects of an experimental eects are evaluated by
comparing two or more subject groups who receive dierent interventions.
Participants who meet inclusion criteria are selected and randomly distributed
between the control and experimental condition. The rationale for random
distribution is to ensure that variations between individuals are mitigated. If
participants were assigned to a condition in a non-random way and had
characteristics in common the results of the dependent variables could be attributed
to these characteristics, and not the intervention. During the experiment, each
268 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
experimental group receives a brief intervention for an equal period of time. Then,
data from each group are analyzed and compared. The between-groups design is
widely used to determine the eectiveness of psychological interventions, but it
has its limits in regards to understanding psychotherapy, as it does not allow for
the assessment of the psychotherapeutic process, such as the regular changes that
occur at each session and the identication of the mechanisms of clinical change
involved (Kazdin, 2003).
In single-case designs, comparisons of the same subject are examined at
dierent time points as opposed to few comparisons in several subjects, as in
cross-subject designs. This design is also called “subject as its own control” or
“n = 1” (Sidman, 1960). Kazdin (2003) describes three minimum specications that
a single-case design must meet in order to be able to infer causality between the
independent variables and the dependent variables:
• Continuous assessment: the most important aspect of this design is that the
dependent variables are recorded continuously, preferably before the
intervention is introduced and while it is presented or removed. This makes
it possible to infer the eect of the independent variable on the dependent
variable and to implement the next requirement.
• Baseline: continuous evaluation allows the experimenter to know the level of
the dependent variables before the intervention and to develop a hypothesis
of how the data would respond if no treatment was applied. The baseline
allows for visualization and comparison of the participant’s repertoire before
and after the administration of the independent variable.
• Stability: it is necessary that the dependent variables show stability in
their tendency and variability, so that it is possible to contrast their
changes after the introduction of independent variable. In order to draw
conclusions about the eect of a psychotherapy, the target behavior rate
must demonstrate a clear trend and maintain an average proportion at the
baseline. This behavior should change, so that it is clear that the participant’s
behavioral pattern has been aected. The more stable the dependent
variables are in the baseline, the easier it will be to infer the impact of the
independent variables.
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 269
FAP and substance abuse
FAP studies historically have a preference for single-subject experimental
designs for two reasons: 1. given the analytic-behavioral tradition, researchers
understand that data must come from comparing the same participant at dierent
times and not between individuals; and 2. single-subject design prioritizes the
description of the behavioral processes underlying the observed changes and allows
the analysis of the therapeutic process over time.
1.3 FAP and substance use disorders (SUD)
Substance abuse and substance dependence are psychopathologies described
in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (APA)
as a problematic pattern of substance use despite the adverse consequences that
consumption brings to the user. A substance dependence diagnosis includes
tolerance, withdrawal symptoms and a pattern of compulsive use, all of which are
not found in an abuse diagnosis. The fth edition of the manual (APA, 2013) unied
the two cases into a single diagnosis of Substance Use Disorder (SUD). The new
denition species issues involving psychoactive substances within a continuum of
mild to more serious consequences. A diagnosis of SUD is more in line with the
behavior analysis tradition, as both interpret psychopathologies as a set of
behaviors in the individual’s repertoire that lead to aversive consequences for
themselves and for third parties (Tourinho & Luna, 2010), modifying the division
between abuse and dependency as arbitrary.
Researchers using a behavioral approach argue that drug use and the client’s
repertoire decits can be conceptualized as learned behaviors. Substance abuse
is maintained by pharmacological and social consequences, which can reinforce
behavior positively or negatively (Higgins, Heil, & Sigmon, 2007). The substance
becomes relevant by providing temporary access to reinforcers, such as fun, friends,
and the elimination of aversive events, that the individual would not be able to
obtain otherwise. However, the contingencies cause serious problems in the long run.
When trying to stop substance use, the users nd themselves without the necessary
skills to deal with increasingly intense diculties, which leads them to relapse. The
cycle of abuse does not confer the opportunity for emotional or behavioral
development, and the cycle becomes self-perpetuating (Banaco & Montan, 2018).
Behavioral therapists hold the rationale that new behaviors must be taught,
so that clients can interact more eectively in their physical and social environment.
270 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
Access to drug-free satisfaction, identication of factors that perpetuate
consumption, and the development of problem-solving skills can end the SUD
pattern and promote a healthier lifestyle (Higgins et al., 2007). Due to behavioral
therapy’s therapeutic achievements, it has been established that these techniques
should be included eective treatment. Thus, there have been many variations of
behavioral therapies proposed to treat SUD (Carroll & Onken, 2005).
Despite advances in the eld of psychology, only a minority of substance
users receiving psychotherapy treatment achieve successful outcomes. This calls
for the development of new psychotherapies and the adaptation of therapies that
have already demonstrated being eective for other disorders. Understanding the
active components of an intervention makes it possible to remove the expendable
elements and maximize the therapeutic eect when applied to a new disorder
(Carroll & Onken, 2005). This is the case with the innovation of third-wave
behavioral therapies, which have shown to be eective for several diagnoses and
now are being used to treat SUD. One of these new approaches includes FAP
(Kohlenberg & Tsai, 1991).
FAP can be understood as an adjunct treatment that creates opportunities
for the therapist to reinforce more eective interpersonal repertoires in-session,
while conducting other interventions for SUD. The expectation is that the improved
in-session repertoires will be generalized to out-of-session contexts, reducing the
aversiveness of social events and enabling the client’s contact with social and
aective reinforcers (Holman et al., 2012), thus, decreasing the likelihood of relapses.
There are a few articles related FAP and SUD in the literature. Despite some
interesting data, researchers are unable to demonstrate the primary eect of FAP
on the performance of participants and a related reduction of drug use. As we will
see, researchers do not use behavior analysis methods and apply dierent
interventions using FAP.
Paul, Marx, and Orsillo (1999) presented a client with problems related to
exhibitionism and marijuana abuse who was treated with Acceptance and
Commitment Therapy (ACT) and FAP strategies. The case formulation revealed
that marijuana use was sustained by predisposing exhibitionist behavior, because it
reduced the aversiveness of its social consequences. The therapeutic goals were to
accept exhibitionist thoughts and feelings, while reducing the frequency of
exhibitionism and substance use and increasing social contact. The authors used
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 271
FAP and substance abuse
self-report measures to record frequency and intensity of impulses, episodes of
exposure, masturbation, and drug use. ACT strategies improved social competence,
but marijuana use decreased only with the introduction of FAP. The hypothesis was
that the reinforcement of CRB2s, exposed aversive content, facilitated the
acceptance of aversive stimulation in session and, when generalized out of session,
helped with the tolerance of urges to expose a precipitator of marijuana use. As for
the study’s limitations, there was no record of the client and therapist behaviors in
session, thus, it was not possible to infer the direct eect of FAP on CRBs and the
intervention mechanism of change.
Giord et al. (2011) compared a medication intervention to medication with
ACT and FAP therapy for cigarette smokers. Objective measurements were used for
tobacco, and standardized inventories were used in order to identify psychiatric
symptoms and behavioral processes. The results pointed to improvements after
the intervention and sustained improvements in the one-year follow-up for the
ACT/FAP group. Despite the randomized trial’s great experimental control, it was
not possible to dene the eect of each therapy and to individually evaluate the
mechanisms involved. Like the Paul et al. (1999) study, there was no record of the
client and therapist behaviors in session.
Holman et al. (2012) conducted a study with ve depressive and nicotine
dependent clients who received a combination of Behavioral Activation, Smoking
Cessation, ACT and FAP procedures. Self-report measures were acquired for
depression and smoking, as well as objective measures for smoking. After the
intervention, four clients experienced remission of the clinical criteria for major
depression, and three were abstinent from tobacco – the other two had a signicant
decrease in consumption. Nevertheless, like the other two studies, the results did
not allow for inference of the eect regarding each intervention type and the
underlying change processes.
Indeed, previous research has employed FAP elements along with other
behavioral therapies and presented promising results regarding the participant’s
drug use. However, the methodologies did not allow researchers to analyze the
isolated eect of FAP and its mechanism of clinical change. The current study
aims to exclusively evaluate the eectiveness of FAP and the essential components
of change on the repertoire of clients with SUD as measured by CRBs and changes
in the pattern of substance abuse. Our hypothesis was that the therapist’s
272 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
contingent responding would be responsible for decreases in the frequency of
CRB1s and increases in CRB2s, which would lead to improvements in drug use
after the intervention.
2. Method
2.1 Participants
The study was approved by the Ethical Research Committee, and it is
registered under the Presentation Certicate of Ethical Appreciation (Certicado de
Apresentação de Apreciação Ética [CAAE]) No. 45003515.2.0000.5561. Participants
provided informed consent. The participants were two clients diagnosed with SUD
who did not have any unstable medical and/or psychiatric conditions. Participant 1,
Rafael, was 63 years old and had originally sought out therapy for alcoholism. The
identied CRB1s included client responses that were dismissive of the therapist’s
knowledge about him, such as supercial talk when relevant topics are evoked and
extreme politeness when judging others. CRB2s included direct verbalizations of
his behaviors/feelings and the contingencies that inuenced him and giving
opinions about people and events. Rafael started using Fluoxetine at 60mg in the
third month of treatment. Participant 2, Lucas, was 36 years old and originally
sought out treatment for cocaine/crack use. His CRB1s included decits of analyzing
contingencies of reinforcement that control his behavior and that aected third
parties. Not understanding what aected him lead him to feelings of shame and
depression (i.e., evaluating himself as “ignorant” for making a mistake). When he
did not understand what inuenced others, especially when someone denied a
request, he got annoyed or upset (i.e., not understand why his wife doesn’t want
him to go out). The CRB2s included eectively relating to events involving other
people (i.e., understanding the behavior of other people) and him (i.e., understanding
his own behavior). When he arrived at the institution, he had been prescribed
Fluoxetine at 20mg.
The researcher-therapist (the rst author), who graduated six years ago, was
a specialist in Clinical Psychology/Behavioral Therapy and had experience in private
practice and institutions for the treatment of SUD. He was supervised in a therapy
group coordinated by another psychologist (the second author) with 13 years of
clinical experience, including expertise in conducting research and training in FAP.
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 273
FAP and substance abuse
Table 2.1.1. Contingencies of reinforcement for participants’ Rafael and
Lucas CRBs.
Antecedent Response Consequence
Rafael’s
CRB1
Therapist inquiries
about feelings/
thoughts or evaluation
about others.
Supercial talk and
extreme politeness.
Therapist stops
questioning.
Rafael’s
CRB2
Therapist inquiries
about feelings/
thoughts or evaluation
about others.
Verbalizations about his
behavior/feelings.
Opinions about people
and events.
Therapist’s social
reinforcement.
Lucas’
CRB1
Therapist inquiries
about an interpersonal
conict.
Decits in analyzing
contingencies that
aected him and third
parties.
Ineective social
response.
No consequences
are provided and he
presented feelings
of sadness and
ange r.
Lucas’
CRB2
Therapist inquiries
about an interpersonal
conict.
Eectively relating to
events about himself and
other people.
Eective social response.
Therapist’s social
reinforcement.
2.2 Environment
Data collection was performed in a specialized clinic for SUD, with a minimum
voluntary hospitalization period of four months. Resocialization of the residents
took place by phone contact (after seven days of hospitalization), family visits
occurred on Saturdays and/or Sundays (after 15 days), and residents had the option
to sleep in their residence on weekends (after 30 days). If they were not discharged
after four months, the clients split their time between participating in the activities
of the institution and looking for a job or professional courses. The treatment
package included weekly medical appointments, SUD lectures, cognitive-behavioral
psychoeducation, group psychoanalytic psychotherapy, alcoholics anonymous
groups, group family counseling, and working out at an external gym. Individual
psychotherapy sessions lasted between 50 and 60 minutes.
Participant 1, Rafael, spent eight months in the clinic, in which during the
rst four months he followed the rehabilitation rules and after the fth month he
274 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
alternated between the institution’s psychotherapies and external activities during
the week, such as buying newspapers, or going to a coee shop, bakery and medical
appointments. When a relapse occurred, the institution’s clinical professionals
decided how long it would be appropriate for Rafael to stay in detoxication without
going out. Participant 2, Lucas, completed treatment two weeks ahead of schedule
and adhered to rehabilitation standards, and he did not go out during weekdays.
2.3 Measurement
• Functional Analytic Psychotherapy Rating Scale (FAPRS) (Callaghan & Follette,
2008): this scale is a system for coding client and therapist behavior in
FAP sessions, transcribed from audio or video recordings, containing in its
original version seven codes for the behavior of the client and 15 for the
therapist. For the current study, client coding categories included codes for
clinically relevant behaviors (CRB1, CRB2, and CRB3), problem or improvement
in behaviors outside therapeutic relationship (O1 and O2), talk that focuses
on the therapeutic relationship (CTR) and general progression of the session
(TPR). Therapist codes included observing CRBs (rule 1), evoking CRBs (rule
2), eectively responding to CRBs (rule 3-1, rule 3-2), asking about the
eect of the intervention on the client’s repertoire (rule 4), generalization
strategies (rule 5), eectively responding to O1 and O2 (TO1 and TO2),
general progression of the session (TPR), talk that focuses on the therapeutic
relationship (TTR), and missing an opportunity to respond to CRBs or
responding ineectively (rule 3-INF). Codes are given to each turn of speech
during the session, thus, after categorization it is possible to quantify the
change in client behavior after the introduction or withdrawal of specic
therapist behaviors.
• Timeline followback method (TLFB) (Sobell & Sobell, 1994): The TLFB measures
the frequency and intensity of the client’s drug use before and after
treatment. A record of the amount ingested is used for alcohol, while, for
other drugs, only use or no-use is recorded, due to the diculty of measuring
the amount consumed. The TLFB can be administered in three formats: pencil-
and-paper, interviewer or computer. In all modalities, an annual calendar is
presented to the participant along with verbal aids about important dates
and events (birthdays, vacations, holidays etc.), so that the participant can
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 275
FAP and substance abuse
recall and report retrospectively their use each day before the treatment, in
a period ranging from seven days to two years.
2.4 Procedure
We conducted a single case quasi-experimental design A/A + B (Kazdin,
2003), in which in the rst phase (A) carried out case conceptualization and
strategies of behavioral-analytic therapy (terapia analítico-comportamental [TAC])
(Tourinho & Luna, 2010) regarding the analysis of external contingencies in the
client’s daily life. FAP (A + B) was introduced after the CRBs were identied for each
participant. In the second phase FAP was incorporated, focusing on evoking and
reinforcing CRBs during sessions.
20 psychotherapy sessions were recorded and transcribed for the rst
participant, while 18 sessions were recorded and transcribed for the second
participant. For each phase, ve sessions were coded for each participant to establish
single case standards for data collection as proposed by Kratochwill et al. (2013).
Client 1’s coded sessions were one, three, six, nine and 12 (phase A) and 13, 14, 15, 16
and 17 (phase A + B). Client 2’s coded sessions were one, three, ve, seven and ten
(phase A), and 11, 12, 13, 14 and 15 (phase A + B). These data were chosen in order to
evaluate CRBs from the rst to the last session for phase A and an ongoing assessment
of FAP outcomes when the therapy was introduced in phase A + B (the next ve
sessions). Three months after treatment, a follow-up session occurred in order to
assess the progress sustained and a second TLFB application, which made it possible
to compare drug consumption before and after the trial. Participant 1 stayed in the
clinic in the period between A + B and follow-up. He continued his treatment and
therapy processes. Participant 2 left the clinic and started therapy with another
professional. He came to the clinic for the follow-up session.
The therapist-researcher conducted all FAPRS coding. A reliability assessment
was carried out between the researcher and independent coders trained in the
FAPRS. A session was randomly selected from each phase for each participant
(sessions six and 16 for participant 1, and 5 and 15 for participant 2), which were
then analyzed by 2 assessors, who also signed a TCLE before starting the process.
Kratochwill et al. (2013) suggested 80% agreement or 0.6 Kappa to validate an
observer’s agreement. Both criteria were adopted in this study. First, the coder
studied the FAPRS guide and case conceptualization for its client. After an individual
276 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
meeting (30 minutes) with the researcher to present the conceptualization, a
categorization practice was conducted with transcription and audio les from the
sessions that were not previously assessed.
After the categorizations were sent, there was an hour-long meeting with
the lead researcher, followed by two therapy sessions with the participants for a
nal assessment. The initial outcomes were not satised by the rst coder
(responsible for participant 1). After the rst therapy session and with a new
categorization, the rst coder achieved an 82.31% agreement and 0.74 Kappa (rst
therapy session) and 76% agreement and 0.64 Kappa (second therapy session).
The second coder (in charge of participant 2) also did not present sucient indexes,
therefore, it was not possible to calculate Kappa, due to the dierences found in
the number of categories that were used by the lead researcher (13) and the
assessor (7). Additionally, there were categories with low occurrences in session
(only one or two), which decreased the agreement probability. Thus, a new coding
scheme was carried out to make possible to calculate Kappa. In the third attempt,
the results were 82.97% agreement and 0.74 Kappa (rst therapy session) and
76.19% agreement (second therapy session). It was not possible to calculate Kappa
for the second session due to the dierences in the number of categories.
As was previously argued by Villas Bôas (2015), the time available for
agreement assessment could aect the data analysis. Training can take up to 100
hours according to some studies, which was not feasible for this work. The main
challenges faced by the independent assessors were to understand FAPRS guide
(change categories such as rule 1 and rule 2 and TTR and TPR), and not an
unawareness of the cases, which could both be addressed with sucient training
time. Thus, the indexes that scored below the established criteria were classied as
reasonably satisfactory. For data analysis, the lead researcher calculated the
percentage of occurrences for each category code, in order to compare sessions
even if they had a dierent total number of verbalizations. All data were analyzed
through visual inspection (Kazdin, 2003; Kratochwill et al., 2013; Sidman, 1960).
3. Results and discussion
Figure 3.1 presents the CRB1s and CRB2s of both participants during the
treatment phases and the follow-up sessions. This shows that the introduction of
FAP led to a lower frequency of CRB1s and a higher frequency of CRB2s for both
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 277
FAP and substance abuse
participants. While the positive results remained stable until the end of sessions,
CRB1s had dierent results for each client. During the FAP sessions, participant 1’s
CRB1s decreased (top panel), but this trend began in phase A, which raises questions
concerning causality. Changes in magnitude (session 12 = 11.91%; session 13 =
3.47%) and changes in stability were considered a FAP eect. In contrast, participant
2’s CRBs were not suddenly altered by FAP.I fact, the percentage after FAP was
equivalent to that of session 12 during phase A (9.42%). Similar to what was
proposed by Villas Bôas (2015), we hypothesized that the dierence found could be
explained by the nature of the CRBs. The therapist focused on a specic repertoire
of establishing more eective verbal relations about events, which could not be
done in only one therapy session. Thus, it was necessary for the client to express
several relations and the therapist to reinforce the desired relations. Our hypothesis
was corroborated by the CRBs that increased in this session (11.23%), the best
result in this phase. Simões (2014) described a similar phenomenon, which
evaluated the impact of metaphors on the client’s verbal relations about events.
Such changes happen only in latter stages after the independent variable has been
introduced.
278 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
Figure 3.1. Percentage of CRB1S and CRB2s (problem behaviors and
therapeutic progress) over ten sessions of psychotherapy for Rafael (upper
panel) and Lucas (lower panel), analyzed according to experimental phases A
= TAC and A + B = FAP and in the follow-up session.
Rafael
TAC FAP
Percentage of CRBs
Number of sessions
1 3 6 9 12 13 14 15 16 17
50
45
40
35
30
25
20
15
10
5
0
CRB1 CRB2
Follow
Lucas
TAC FAP
Percentage of CRBs
Number of sessions
1 3 5 7 10 11 12 13 14 15
50
45
40
35
30
25
20
15
10
5
0
Follow
CRB1 CRB2
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 279
FAP and substance abuse
Follow-up results for participant 1, who continued the treatment in group
and individual sessions, suggest that therapeutic progress continued (CRB1 = 7.87%
and CRB2 = 12.82%). It could be argued that the continuity of the treatment
contributed to the improvements in CRB2s, in addition to the increase in CRB1s
when compared with session 17. A session analysis revealed that some of the
behaviors that could be classied as CRB1s (1.74%) varied in quality once the client
openly stated that he “would not like to talk about a certain matter” instead of
being quiet or briey speaking about it. Such a result revealed even more
improvement.
Participant 2’s results were not the same. He did not nish the 20 planned
sessions, and only attended 18. He did not keep in touch with the researchers and
started a dierent psychotherapy with another professional. The follow-up session
revealed that his CRB1s remained the same (4.66%) and CRB2s were the same as
before the intervention (1.69%). It is possible that the short period of time did not
allow the CRB2s to be established and generalized to out-of-session relationships
(the early termination was motivated by a problem behavior). The CRB1s could be
explained by the fact that the therapist focused on managing relapses, which
decreased the verbalizations about relations between interpersonal events. Even in
smaller proportions, the CRB1s which were analyzed evoked the same relationship
challenges that can raise the probability of drug use.
The results of this study can be seen as a systematic replication (Sidman,
1960) of previous studies, including quasi-experimental designs (Villas Bôas,
2015), single case reversion experimental designs (Geremias, 2014), and multiple
baseline between participants designs (Lima, 2017). In the current study, during
the FAP therapy sessions there was a decrease in CRB1s and an increase in CRB2s.
Replication of this nding across the two clients also strengthened the hypothesis
that the same active principle most likely inuenced the repertoires of both
participants.
280 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
Figure 3.2. Percentage of rules 3-1 and rules 3-2 (respond eectively to
CRB1s and CRB2s) over ten sessions of psychotherapy for Rafael (upper
panel) and Lucas (lower panel), analyzed according to experimental phases
A = TAC and A + B = FAP.
TAC FAP
Rule3-1 Ru l e3-2
Rafael
TAC FAP
Percentage
Number of sessions
1 3 6 9 12 13 14 15 16 17
50
45
40
35
30
25
20
15
10
5
0
Rule3-1 Ru l e3-2
Lucas
Percentage
Number of sessions
1 3 5 7 10 11 12 13 14 15
50
45
40
35
30
25
20
15
10
5
0
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 281
FAP and substance abuse
Figure 3.2 shows the percentage of eective therapist responses to CRB1s
and CRB2s. Rules 3-1 oscillated for both participants during phase A, and then
increased and remained the same in phase A + B. Therefore, the therapist’s
intervention did not inuence the frequency of client behaviors. Although, adding
rule 3-2 during the second stage reversed the proportions of CRBs. Rule 3-2 was
observed at higher rates when compared to rule 3-1 in all sessions of phase A + B,
except for participant 1’s 14th session (rule 3-1 = 4.36% and rule 3-2 = 3.27%).
Despite the therapist focusing on responding to CRB1s during 14th session, CRB2s
occurred more frequently.
These data contribute to previous FAP studies regarding the eectiveness of
FAP and the therapeutic progress. In highly controlled designs, despite variations in
the proportion of rules 3-1 in FAP and no-FAP phases (Geremias, 2014), rule 3-2
occurred more frequently in FAP therapy sessions and was related to a decrease in
CRB1s and an increase in CRB2s (Kanter et al., 2017). This suggests that the
mechanism of clinical change in this psychotherapy is the therapist’s contingent
response to the client’s improvements rather than problems.
Additionally, Figure 3.3 presents data on drug abuse among participants.
Before the intervention, participant 1 drank daily, with greater frequency during the
week (average of 12 doses) and lower frequency at weekends (average of three
doses). He began the rst treatment (without resocialization) and did not drink,
but when he nished, he drank during the next two days (four and ve doses). Data
was collected in the second institution where he was hospitalized. A relapse not
recorded by TLFB occurred in the week of session 18 and he was instructed not to
leave the clinic for 15 days. On the third day of external activities, he had a second
relapse (ve doses) and was suspended for another 15 days. After this hiatus it was
determined by the technical team that the client could perform his external
activities even if there were relapses. One week later he had a new relapse (eight
doses) and two months later, the last relapse was recorded (three doses).
Participant 2 previously consumed cocaine/crack during the weekdays but
on weekends was able to control himself and spend time with his family. However,
after three consecutive weekends using crack and with the perception that he had
lost control, he sought out the institution. When he was hospitalized, he stayed for
one month without using drugs, started again a few times a week and, then, in an
attempt to prevent relapses, began using every 15 days on average.
282 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
Figure 3.3. Consumption of psychoactive substance three months before
and three months after the quasi-experimental design. Rafael (upper
panel) reported alcohol (doses) and Lucas (lower panel), cocaine/crack (use
and non-use).
Both clients showed quantitative improvements in drug use, and a qualitative
analysis revealed a bigger improvement for Rafael and a smaller one for Lucas.
Participant 1 started a romantic relationship, improved his dialogue with the clinic
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 283
FAP and substance abuse
professionals, and his CRB2s were more improved on the follow-up session.
Probably, the established repertoire and positive reinforcements that he had
practiced helped him to consume less alcohol. On the other hand, participant 2 left
the facility before the planned date and did not show the same progress at the
follow-up session. Furthermore, participant 2 suered a sports injury during his
rst month out of the institution, which obligated him to stay at home with his
wife and led to more interpersonal conicts. His drug use as per the TLFB followed
the rate of positive reinforcers he had access to (during the rst month) and
increased with the exposure to negative reinforcers (the second and third months).
If the progress went as expected (improved CRB2s generalized to relationships out
of sessions), he might have experienced fewer conicts and his crack use might
have decreased.
We can identify three main limitations to this work. The rst concerns the
chosen method. The quasi-experimental design revealed a correlation between the
independent variable and the dependent variable, but we cannot be sure about this
relationship. It is necessary to use more powerful designs that would allow
researchers to measure the inuence of FAP on CRBs and participants’ drug use. We
suggest a reversion design or multiple baseline (Kazdin, 2003; Kratochwill et al.,
2013). Despite the relative success with the agreement assessment compared to
the training time, we question if the results could have been better if the assessor
had a better understanding of the FAPRS. We recommend a longer training time
(Kratochwill et al., 2013), which could then improve the study’s internal validity.
Finally, the record of drug use does not allow for an establishment of a clear relation
between FAP and drug use. Thus, we suggest that a more objective measure (for
instance, urine tests) and daily/weekly (instead of pre and post-test) measures
allow for a more eective evaluation of how CRB changes increase or decrease
in drug use.
4. Final remarks
Previous studies of FAP and SUD have demonstrated improvements in
participants’ behavioral repertoire and drug use. Nevertheless, they do not
specically point out to individual eects of FAP and its mechanism of change. This
quasi-experimental design made it possible to assess the impact of FAP, and to
identify changes in the interpersonal repertoire for both participants, which resulted
284 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
from the therapist’s contingent responding to behavioral improvements. However,
the data shows that participant 1, who had a longer intervention process, had
better outcomes from therapy. Future researchers should apply single-case designs
with more precision, sucient training time for agreement assessor, and more
straightforward and continuous measures to assess drug consumption.
References
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental
disorders (5th ed.). Washington: American Psychiatric Association. doi:10.1176/appi.
books.9780890425596
Banaco, R. A., & Montan, R. N. M. (2018). Terapia analítico-comportamental. In N. A.
Zanelatto, & R. Laranjeira (Orgs.). O tratamento da dependência química e as terapias
cognitivo-comportamentais: Um guia para terapeutas (2nd ed.) (pp. 115–132). Porto
Alegre: Artmed.
Callaghan, G. M., & Follette, W. C. (2008). FAPRS manual: Manual for the functional
analytic psychotherapy rating scale. The Behavior Analyst Today, 9(1), 57–97
doi:10.1037/h0100648
Carroll, K. M., & Onken, L. S. (2005). Behavioral therapies for drug abuse. The American
Journal of Psychiatry, 168(8), 1452–1460 doi:10.1176/appi.ajp.162.8.1452
Carvalho Neto, M. B. (2002). Análise do comportamento: Behaviorismo radical, análise
experimental do comportamento e análise aplicada do comportamento. Interação
em Psicologia, 6(1), 13–18. doi:10.5380/psi.v6i1.3188
Giord, E. V., Kohlenberg, B. S., Hayes, S. C., Pierson H. M., Piasecki, M. P., Antonnucio,
D. O., & Palm, K. M. (2011). Does acceptance and relationship focused behavior
therapy contribute to bupropion outcomes? A randomized controlled trial of
functional analytic psychotherapy and acceptance and commitment therapy for
smoking cessation. Behavior Therapy, 42, 700–715. doi:10.1016/j.beth.2011.03.002
Geremias, M. C. G. (2014). Manejo de esquivas emocionais na psicoterapia analítica
funcional: Delineamento experimental de caso único. (Unpublished Master’s thesis).
Universidade de São Paulo, São Paulo. doi:10.11606/D.47.2014.tde-17032015-
150030
Higgins, S. T., Heil, S. H., & Sigmon, S. C. (2007). A behavioral approach to the treatment
of substance use disorders. In P. Sturmey (Ed.), Functional analysis in clinical treatment
(pp. 261–282). Burlington: Elsevier. doi:10.1016/B978-012372544-8/50015-X
Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286 285
FAP and substance abuse
Holman, G., Kohlenberg, R. J., Tsai, M., Haworth, K., Jacobson, E., & Liu, S. (2012).
Functional analytic psychotherapy is a framework for implementing evidence-
based practices: The example of integrated smoking cessation and depression
treatment. International Journal of Behavioral Consultation and Therapy, 7, 58–62.
doi:10.1037/h0100938
Kanter, J. W., Manbeck, K., Kuczynski, A. M., Maitland, D. W. M., Villas Bôas, A., & Reyes,
M. (2017). A comprehensive review of research on Functional Analytic Psychotherapy.
Clinical Psychology Review, 58, 141–156. doi:10.1016/j.cpr.2017.09.01012
Kazdin, A. E. (2003). Research design in clinical psychology (4th ed.). Needham Heights:
Ally & Bacon.
Kohlenberg, R. J., & Tsai, M. (1991). Functional analytic psychotherapy: A guide for creating
intense and curative therapeutic relationships. New York: Plenum. doi:10.1007/
978-0-387-70855-3
Kratochwill, T. R., Hitchcock, J. H., Horner, R. H., Levin, J. R., Odom, S. L., Rindskopf, D.
M., & Shadish, W. R. (2013). Single-case intervention research design standards.
Remedial and Special Education, 34(1), 26–38. doi:10.1177/0741932512452794
Lima, G. O. (2017). Psicoterapia analítica funcional como tratamento do transtorno de estresse
pós-traumático: Delineamento experimental de caso único. (Unpublished Master’s
thesis). Universidade de São Paulo, São Paulo. doi:10.11606/D.47.2018.tde-
06022018-162328
Paul, R. H., Marx, B. P., & Orsillo, S. M. (1999). Acceptance-based psychotherapy in the
treatment of an adjudicated exhibitionist: A case example. Behavior Therapy, 30,
149–162. doi:10.1016/S0005-7894(99)80051-3
Sidman, M. (1960). Tactics of scientic research: Evaluating experimental data in psychology.
New York: Basic Books.
Simões, E. F. (2014). Manejo de metáforas em psicoterapia analítico-comportamental.
(Unpublished Master’s thesis). Universidade de São Paulo, São Paulo. doi:
10.11606/D.47.2014.tde-17032015-152454
Sobell, L. C., & Sobell, M. B. (1994). Timeline followback (TLFB) user’s manual. Toronto:
Addiction Research Foundation.
Tourinho, E. Z., & Luna, S. V. (Orgs.). (2010). Análise do comportamento: Investigações
históricas, conceituais e aplicadas. São Paulo: Roca.
286 Psicologia: Teoria e Prática, 22(3), 263-286. São Paulo, SP, set.-dez. 2020. ISSN 1980-6906 (on-line).
doi:10.5935/1980-6906/psicologia.v22n3p263-286
Alan S. Aranha, Claudia K. B. Oshiro, Elliot C. Wallace
Villas Bôas, A. (2015). Efeitos de análises de contingências sobre comportamentos clinicamente
relevantes e sobre mudanças extra sessão. (Unpublished Doctoral dissertation).
Universidade de São Paulo, São Paulo. doi:10.11606/T.47.2015.tde-24092015-
154310
Authors notes
Alan S. Aranha, Department of Clinical Psychology (PSC), University of São Paulo (USP);
Claudia K. B. Oshiro, Department of Clinical Psychology (PSC), University of São Paulo
(USP); Elliot C. Wallace, Center for the Science of Social Connection, Center for the
Study of Health and Risk Behavior (CSSC/CSHR), University of Washington School of
Medicine (UW).
Correspondence concerning this article should be addressed to Alan Souza Aranha,
Avenida Prof. Mello Moraes, 1721, Butantã, Cidade Universitária, São Paulo, SP. CEP
05508-030.
E-mail: alansaranha@gmail.com