Journal of Fluency Disorders 32 (2007) 95–120

Stages of change and stuttering: A preliminary view

Jennifer Floyd a , Patricia M. Zebrowski b,∗,1 , Gregory A. Flamme c
a 2392 E. 116th Ct., Thornton, CO 80233, 303-920-0471, United States
b Department of Speech Pathology Audiology, University of Iowa, 127C SHC,
Iowa City, IA 52245, United States
c Department of Speech Pathology Audiology, Western Michigan University,

Kalamazoo, MI 49008, United States

Received 22 December 2006; received in revised form 7 March 2007; accepted 9 March 2007

Abstract
As a way to better understand the process of change that occurs in stuttering, Craig [Craig, A. (1998).
Relapse following treatment for stuttering: a critical review and correlative data. Journal of Fluency
Disorders, 23, 1–30] compared the behavioral changes that people who stutter often experience with and
without treatment to those that have been observed for certain (non)addictive behavior disorders such as
smoking, overeating, phobia and anxiety disorder. The process underlying these behavioral changes has
been described by the transtheoretical or “stages of change” model, which is a model of behavior change
that can illuminate “where” a person is in the process of change, and how this may relate to the outcome of
either treatment or self-change attempts [Prochaska, J. O., & DiClemente, C. C. (1986). The transtheoretical
approach. In J. C. Norcross (Ed.), Handbook of eclectic psychotherapy. New York: Brunner/Mazel]. The
purpose of the present study was to analyze the extent to which the responses of adults who stutter on a
modified Stages of Change Questionnaire yield interrelations among questionnaire items that are consistent
with a stage-based interpretation. Results of both confirmatory and exploratory factor analyses indicated
that while the modified questionnaire was a relatively good fit for participant responses, the structure
derived from the exploratory analysis provided a significantly better fit to the observed data. Results suggest
that a questionnaire incorporating items that better reflect the unique behavioral, cognitive and affective
variables that characterize stuttering may better discriminate stages of change in people who stutter as they
move through therapy, or are engaged in self-directed change.

Educational objectives: After reading this paper, the learner will be able to: (1) describe the transtheoretical
or “stages of change” model; (2) describe the various processes that are associated with different stages of

∗ Corresponding author. Tel.: +1 319 3358735; fax: +1 319 3358851.

E-mail addresses: [email protected] (J. Floyd), [email protected] (P.M. Zebrowski),
[email protected] (G.A. Flamme).
1 Tel.: +1 303 920 0471.

0094-730X/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jfludis.2007.03.001
96 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

change; (3) summarize research findings in stages of change as they apply to a variety of clinical populations;
(4) discuss the applicability of the findings from the present study to stuttering treatment, and (5) relate
conventional strategies and techniques used in stuttering therapy to different stages in the process of change.
© 2007 Elsevier Inc. All rights reserved.

Keywords: Stages; Change; Stuttering

In an attempt to explain the different factors that contribute to treatment responsiveness in

stuttering therapy, Zebrowski and Conture (1998) identified both “client independent” and “client
dependent” variables that may affect treatment outcome. So-called client independent factors
include observable and measurable stuttering behaviors, such as the frequency, duration, and
type of speech disfluency an individual produces, as well as environmental factors that might
influence the frequency, severity and variability of stuttering. Client dependent variables, on the
other hand, are comprised of those affective and cognitive factors or characteristics that have
been implicated in the development of intractable stuttering. For example, a number of studies
of stuttering treatment have shown that variables such as avoidance, external locus of control,
production of learned compensatory behaviors, negative attitudes about speech, and high levels
of trait anxiety are associated with poorer long-term gains from therapy as well as relapse (e.g.,
Blood, 1993; Craig, 1998; Guitar, 1976, 1998; Guitar & Bass, 1978). In a more recent study, Huinck
et al. (2006) showed that a pretreatment profile consisting of measures of stuttering severity and
severity of both negative emotions and cognitions yielded subgroups that experienced different
treatment outcomes. Finally, a number of authors have argued that the client’s motivation and
willingness to change have a critical impact on stuttering treatment outcome. In particular, the
client’s readiness for change as it relates to the timing of therapy enrollment is an important factor
in success; that is, beginning therapy when one is most ready for change leads to a more positive
outcome and maintenance of therapy goals (Blood, 1993; Manning, 2001, 2006; Shapiro, 1999).
In 1998, Craig published an excellent review and extension of the research in stuttering treat-
ment outcome, with a special focus on the factors that may predict relapse. He concluded that
there are limited factors that alone can reliably predict relapse; rather a multifactorial model of
relapse has the most validity. Craig further noted that relapse in stuttering therapy is relatively
typical, and that it is strikingly similar to the “relapse-success cycle” (p. 23) that characterizes the
change process in therapy for other behavioral and psychological disorders and addictions (e.g.,
smoking, eating disorders, phobia, etc.). For example, the cyclical nature of the smoking cessation
process, with its alternating periods of improvement and relapse over time, also characterizes the
movement toward long-term change in speech fluency that individuals who stutter experience over
the course of therapy. Manning (2001, 2006) echoed this point by describing stuttering therapy
as a “pattern of cyclical movement through a process of gradually more successful management
of stuttering-related behaviors and processes” (2006; p. 140). Until the new behaviors the client
is learning in therapy become stable and predictable, it is likely that he or she will move back and
forth between periods of more and less fluency on the path to recovery, at times even reverting
back to pre-therapy levels of stuttering (or “relapse”).
Like Craig, Manning noted that the pattern of change and relapse seen in stuttering therapy
is very similar to that observed in other disorders, and suggested that previous research in the
broader area of change (e.g., Prochaska, DiClemente, & Norcross, 1992), would be a good place
to start to examine the change process in the stuttering population, with an eye toward developing
effective anti-relapse strategies.
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 97

1. Theories of change: the theory of planned behavior and the transtheoretical or

stages of change model

Research in the fields of health and sports psychology, and psychotherapy, has long been
devoted to the study of how people make durable behavioral change. Over the years, a number
of theories of change have been developed that have provided the framework for intervention
approaches designed to help people acquire positive health behaviors (e.g., smoking cessation)
and optimal physical performance. In general, an individual’s beliefs and intentions have been the
cornerstone of arguably the two dominant models of health behavior change: the theory of planned
behavior (Ajzen, 1985) and the transtheoretical or “stages of change” model (Armitage, 2006;
Prochaska & DiClemente, 1984). Briefly, the major premise of the theory of planned behavior
is that an individual’s decisions about behavior change are rooted in predetermined intentions,
and it is these intentions that are the best predictors of what people will do (Sarafino, 2006).
Intention to change is a single construct that is based on a summary of three main judgment
factors, namely the person’s perceptions of (1) how acceptable or unacceptable the behavior is
to him or her (attitude), (2) how acceptable or unacceptable the behavior is to others (subjec-
tive norm), and (3) the likelihood of success in changing or acquiring behavior (self-efficacy).
The major focus of the theory of planned behavior is to predict and understand health-related
change; thus researchers investigating this model typically use prospective designs (Armitage,
2006).
The transtheoretical or stages of change model is an integrated model that includes cogni-
tive, affective and behavioral constructs that are described in other well-established theories of
intentional change, including the theory of planned behavior (Ajzen, 1985). The transtheoreti-
cal model is based on the observation that successful change can be demonstrated through both
self-change and “professionally assisted” approaches, suggesting that there are “basic, common
principles that can reveal the structure underlying change occurring with and without (psycho)
therapy” (Prochaska et al., 1992, p. 1102). It is a model of intentional change, in that it attempts to
explain how decision-making, rather than sociological or biological influences, impacts behavior
change. The major focus of the transtheoretical model is to understand how people change, or
the process (not solely the outcome) of intentions to change. Most notably, the model attempts
to understand change readiness as an important piece of the entire process of behavioral change.
In a series of studies, Prochaska and co-workers developed an instrument that could both delin-
eate and validate the core elements, or stages of change (e.g., McConnaughy, Prochaska, &
Velicer, 1983). Using responses to a 32-item scale administered to a wide range of individuals
across various clinical populations (i.e., psychotherapy clients, people in smoking cessation and
weight-loss programs, etc.), Prochaska and co-workers employed a principal components anal-
ysis that yielded six distinct stages of change. These stages were defined as: precontemplation,
contemplation, preparation, action, maintenance, and termination. Additional analysis of partic-
ipant responses demonstrated that adjacent stages (e.g., precontemplation and contemplation)
were more highly correlated than nonadjacent stages (e.g., precontemplation and maintenance),
suggesting that as people change, they tend to move through a predictable sequence of behavior
change stages. Finally, subsequent work demonstrated that there are specific levels and pro-
cesses of change that are associated with each stage, and that therapy is most successful when
processes are “matched” to their related stages (Prochaska, 1999; Prochaska & DiClemente,
1984).
Since its introduction, the stages of change model has been applied in many settings to individu-
als exhibiting a variety of behaviors requiring change. Numerous published articles have confirmed
98 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

the relevance of the model to the change process inherent in such areas as smoking cessation,
cocaine cessation, weight control, adoption of a low-fat diet and exercise, elimination of adolescent
delinquent behavior, safer sex practices and condom use, sunscreen use, increasing frequency of
mammography screening, and treatment of a variety of eating disorders (e.g., DiClemente, 1993;
Marcus & Simkin, 1994; Pallonen et al., 1994; Prochaska et al., 1994; Ruggiero, 1998; Suris,
Carmen Trapp, DiClemente, & Cousins, 1998). Further, the stages of change model of inten-
tional change can illuminate relationships between ‘where’ an individual is in the process of
changing behavior, and the outcome of either therapy or self-change strategies. That is, being in
a particular “stage” of behavioral change can facilitate progress in therapy if there is a “match”
between stage and treatment approach, or present an obstacle to change if there is a mismatch
between the client’s readiness for change and the treatment approach or specific techniques used
in therapy.

2. Understanding change in stuttering

As previously discussed, being able to describe the range of patterns of intentional change
in people who stutter (with or without therapy) can inform the development of treatment in
key ways. The stages of change model seem most appropriate for understanding the change
process in stuttering for a number of reasons. First, as Craig (1998) and Manning (2001, 2006)
have observed, similar to other behavioral change processes, change in stuttering is cyclical and
relapse appears to be typical. The stages model describes change as circular rather than linear,
in that instead of relapse occurring in the final stage of a set sequence of stages, it can, and
usually does, occur at any time over the course of the change process. And, rather than being
viewed as “the end of the line”, the stages model views relapse as a bridge back to an earlier stage
in which the individual has the opportunity to strengthen the skills necessary to make change
more durable and relapse less likely to occur (Brownell, Marlatt, Lichtenstein, & Wilson, 1986).
As Craig argued, understanding this pattern in people who stutter can lead to a comprehensive
model of relapse in stuttering and the subsequent development of effective anti-relapse strategies.
Second, as Manning and others have argued, assessing an individual’s readiness for change seems
particularly important when making decisions about when to enroll in formal treatment, and
what strategies to employ, and the stages of change model allows for the assessment of this
phenomenon.
Recently, Turnbull (2000) described the clinical applicability of the stages of change model to
stuttering treatment by using case examples to argue that the model captures the change process
typically brought about by therapy. She contended that an appreciation of client stage at a particular
point in time would assist clinicians in adjusting treatment levels and processes to meet the
individual’s needs. Turnbull further illustrated her point by discussing commonly used stuttering
therapy strategies as they might best be applied within specific stages of change. As previously
described, the stage of change model has been successfully applied to a variety of disorders that
bear a striking resemblance to developmental stuttering, in terms of the cognitive and affective
layers that constitute the disorder as a whole. These similarities suggest that the stages of change
model can be useful as a method for determining what stage in the change process a person who
stutters is in when entering therapy, and how to select the therapeutic focus that will increase the
probability of a better therapeutic outcome.
An obvious first step in any attempt to apply the stages of change model to either therapeutic or
self-change in adults who stutter is to determine whether the model can be applied to the stuttering
population. With this initial aim in mind, the purpose of the present study was to analyze the extent
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 99

to which the responses of adults who stutter on a modified Stages of Change Questionnaire yield
interrelations among questionnaire items that are consistent with a stage-based interpretation. The
observation of such a factor structure could then serve as the basis for the further development
of criteria to place an individual in any one of the stages of change that have been observed to
characterize other clinical populations. Further, if the therapeutic change process for individuals
who stutter can be described by a stage-based model, the development of optimal treatment
strategies across various stages is possible. Alternately, if the relationships among questionnaire
items are inconsistent with the stages of change model, further work will be necessary to determine
the best way to conceptualize an individual’s status within the process of change. Overall, the
long-term goals of this research are to develop a clinically useful tool for placing an individual
who stutters into a particular stage when therapy commences, and to examine the extent to which
the processes of change described by Prochaska and co-workers are relevant to these stages as
observed for people who stutter.
In order to provide the context for the study discussed in this paper, the following section
presents a more in-depth description of the stages, processes and levels that constitute the “stages
of change” model, along with the questionnaire format for determining an individual’s stage of
change.

3. A brief description of the stages of change model: stages, processes and levels

3.1. Stages

As previously discussed, in the stages of change model, individuals move through five stages
while changing behavior: precontemplation, contemplation, preparation, action, and maintenance.
Each stage represents a different point in the “readiness” of an individual for change, and assumes
both a period of time and a set of tasks that need to be accomplished in order to move to the next
stage (DiClemente, 1993).

3.1.1. Precontemplation
Briefly, in the precontemplation stage, individuals do not recognize a need for change. People
may be in the precontemplation stage because they are denying that a problem exists, or they
may have had prior therapy experiences that have been unsuccessful. They may underestimate
the impact that change will have on their lives, and overestimate the cost involved in changing
(Prochaska, 1999).

3.1.2. Contemplation
Prochaska (1999) described contemplation as the stage in which people “intend to change
in the next six months” (p. 229). They are still ambivalent about the cost/benefit ratio, which
can lead to procrastination. Regardless, individuals who are in contemplation are not ready for
therapy that involves active behavioral change (e.g., treatment that focuses on behavioral speech
changes).

3.1.3. Preparation
In this stage, people have begun taking some form of action to make changes in their behav-
ior. This may take the form of so-called “bibliotherapy” (including web-surfing), obtaining the
names and contact information of relevant helping professionals, or attempting some sort of
self-therapy.
100 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

3.1.4. Action
The action stage is characterized by tangible and deliberate changes at some level. Sometimes
these modifications can be observed and even measured (e.g., changes in speech fluency and/or
stuttering behavior; changes in attitudes about talking and stuttering), and sometimes they cannot
(e.g., retaining a focus in the present when starting to feel anxious about talking, or trying,
successfully or not, to reduce the amount of “overthinking” listener reactions to stuttering).

3.1.5. Maintenance
This stage involves active attempts to prevent relapse or regression. Typically, the individual
does not practice change processes as frequently as he or she did during the action stage. Mainte-
nance, which can be thought of as generalization over time, requires sustained effort or attention
over a prolonged period. One of the primary reasons that people fail to maintain behavior, or
relapse, is that they are not prepared to do “what it takes” to achieve lasting change.
As previously described, the stages of change model assume that change is not an all-or-none-
phenomenon, but rather a gradual movement through stages. The model assumes that regardless
of the change to be made, not all people are ready for the same type of change, and because of this
reality, individuals enter therapy at different stages (Prochaska et al., 1992). Moreover, movement
through the stages requires the use of processes that are appropriate for each stage. In general,
movement is not linear, but spiral, and most people experience relapse and regression to an earlier
stage (Prochaska et al., 1992). When this occurs, however, the individual does not recycle through
the stages endlessly, but instead uses what he or she has learned from prior mistakes, and tends
to move through each stage more quickly the “second time around.”

3.2. Processes

Prochaska and DiClemente’s (1986) analysis of different psychotherapeutic approaches

revealed 10 distinct processes that people use as they move through the stages of change. Table 1
provides descriptions of these 10 processes along with their associated stages of change. In gen-
eral, processes of change represent the activities a person engages in to modify a behavior, affect,
or cognition, and include such actions as consciousness raising, stimulus control and developing
helping relationships with others. These processes appear to be tied to an individual stage, or to
overlap adjacent stages, in that a specific process tends to be used more often when an individual
is in a particular stage of change. For example, the process of “consciousness raising” appears to
be used most often when an individual is in the precontemplation stage, and continues to be used
frequently as he/she moves into contemplation. Consciousness raising includes such activities as
information gathering about oneself and the problem (e.g., through reading and talking to others,
including a professional), making observations and so forth. “Self-liberation”, on the other hand,
is a process used most frequently in the later stage of preparation, and continues to be used as
the individual moves from preparation to action. Self-liberation includes those specific strategies
that people use to commit to change (Prochaska et al., 1992), such as personal and interpersonal
contracts and resolutions.

3.3. Levels

Finally, levels of change represent the components of a particular problem that need to be
addressed in therapy. According to Prochaska and DiClemente (1986) there are five distinct
levels within a particular problem, organized in a hierarchical fashion. From lowest to highest
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 101

Table 1
Stages of change and their associated processes
Stage Processes

Precontemplation (1) Assessing how one feels and thinks about oneself with respect to a problem: value
clarification, imagery, corrective emotional experience
Contemplation (1) Increasing information about self and problem: observations, confrontations, interpretations,
bibliotherapy
(2) Assessing how one feels and thinks about oneself with respect to a problem: value
clarification, imagery, corrective emotional experience
Preparation (1) Assessing how one feels and thinks about oneself with respect to a problem: value
clarification, imagery, corrective emotional experience
(2) Choosing and commitment to act or belief in ability to change: decision-making therapy,
New Year’s resolutions, logotherapy techniques, commitment enhancing techniques
Action (1) Choosing and commitment to act or belief in ability to change: decision-making therapy,
New Year’s resolutions, logotherapy techniques, commitment enhancing techniques
(2) Substituting alternatives for problem behaviors: relaxation, desensitization, assertion,
positive self-statements
(3) Avoiding or countering stimuli that elicit problem behaviors: restructuring one’s
environment (e.g., removing alcohol or fattening foods), avoiding high risk cues, fading
techniques
(4) Being open and trusting about problems with someone who cares: therapeutic alliance,
social support, self-help groups
Maintenance (1) Substituting alternatives for problem behaviors: relaxation, desensitization, assertion,
positive self-statements
(2) Increasing alternatives for nonproblem behaviors available in society: advocating for rights
of repressed, empowering, policy interventions
(3) Rewarding one’s self or being rewarded by others for making changes: contingency
contracts, overt and covert reinforcement, self-reward

Source: Prochaska, J. O., Diclemente, C. C., & Norcross, J. C. (1992). In search of how people change: Applications to
the cessation of smoking. Journal of Consulting and Clinical Psychology, 56(4), 520–528.

points (i.e., least to most complex and/or challenging), these include: symptom or situational,
maladaptive cognitions, current interpersonal conflicts, family or system conflicts and intrap-
ersonal conflicts. For the most part, the stages of change model proposes that clinicians begin
work at the symptom or situational level, because it is a more conscious level and tends to rep-
resent the client’s primary reason for entering therapy. As such, change tends to occur faster
at the symptom or situational level. In stuttering therapy, for example, the main focus at the
symptom or situational level would likely be speech, so that strategies targeting either fluency
(i.e., fluency shaping) or stuttering (i.e., stuttering modification) should yield positive results in a
relatively short time. As the client moves up the hierarchy of levels (i.e., from the symptom to fam-
ily/system and intrapersonal conflicts), those aspects or “layers” of the problem that are related to
sense of self tend to be discovered. These higher levels typically require longer and more complex
therapy.
The levels of change are interrelated, such that changes at one level will lead to change at other
levels. In addition, a particular problem may be maintained at several levels, requiring a clinician
to be prepared to work at any of the five levels of change, individually or simultaneously. Most
importantly, it is thought to be essential to the process of change that the client and clinician agree
on the level at which to work (accommodating the client’s theory of change.)
102 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

3.4. Stages of Change Questionnaire

In 1983, McConnaughy, Prochaska, and Velicer published the first study designed to develop
an instrument for operationally defining the five stages of change. From this investigation emerged
the “Stages of Change Questionnaire”, which they distributed to 155 outpatients across a variety
of mental health settings. The questionnaire, or scale, consisted of 32 statements and employed
a five point Likert format (“1” corresponding to “strongly agree”; “5” corresponded to “strongly
disagree”). A principal components analysis yielded four distinct stages of change: precon-
templation, contemplation, action and maintenance. Results indicated that preparation did not
emerge as a distinct stage, leading the authors to speculate that this stage may be transitory
during relatively fast decision making, and that it involves both contemplation and action (its
two adjacent stages). Additional analysis demonstrated that adjacent stages (e.g., precontem-
plation and contemplation; contemplation and action, etc.) were more highly correlated with
each other than they were with nonadjacent stages (e.g., precontemplation and maintenance),
suggesting that as people change, they move through a predictable sequence. McConnaughy
et al. (1983) concluded that the process of change is best characterized by these four adjacent
stages. This work resulted in a modification of the original questionnaire into a version that con-
tained the original 32 items contained within four, eight-item subscales, each related to one of
the four distinct stages of change: precontemplation, contemplation, action, and maintenance. In
recent studies and clinical papers, the Stages of Change Questionnaire has also been referred to
as the University of Rhode Island Change Assessment (URICA) (e.g., DiClemente & Hughes,
1990; Pantalon & Swanson, 2003; www.uri.edu/research/cprc/Measures/Smoking04urica.
htm).

4. Summary

The stages of change or transtheoretical model has been speculated to be valid for both the
delineation of readiness for change and the description of the change process for people who stutter.
Currently, however, there are no data to either support or refute this hypothesis. The purpose of
this preliminary study, then, was to assess the validity of a stages model to characterize the
change process in the population of adults who stutter, and specifically, to examine the extent
to which responses to items on a modified version of the Stages of Change Questionnaire (or
URICA) would yield a factor structure similar to that which has been observed for other clinical
populations. If so, this factor structure could form the basis for the development of criteria to place
an individual in any one of the stages of change that have been observed to characterize other
clinical populations. Alternately, if the relationships among questionnaire items are inconsistent
with the stages of change model, further work will be necessary to determine the best way to
conceptualize an individual’s status within the process of change.

5. Method

5.1. Participants

Participants for this study were 44 adolescents and adults who stutter between the ages of 16 and
61, with a mean age of 34.9 years. Ten participants were female and 34 were male. Participants were
volunteers recruited through speech-language pathologists and through the National Stuttering
Association (NSA). All participants reported a history of stuttering therapy.
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 103

Individual speech-language pathologists and NSA chapters (hereafter referred to as “gate-

keepers”) were contacted by the investigators, and provided with a brief description of the study
methods. Gatekeepers were asked to provide this description, at their own discretion, to persons
who stutter whom they knew personally, or for whom they currently provided treatment (in the
case of speech-language pathologists). Neither the purpose of the study, nor its related hypotheses,
were shared with either the gatekeepers or potential participants. If and when interested parties
expressed interest, they were given a letter describing the study methods in more detail. Subse-
quently, all interested potential participants were sent a letter describing the study methods again,
and what their involvement in the study would involve. Once an individual agreed to participate,
he or she was given a consent form, and then a modified Stages of Change Questionnaire to com-
plete and return to the first author within a predetermined period of time. The study participants
completed the scale at their own pace and returned it to the first author in a stamped envelope
provided by the investigators.

5.2. Procedure

A modified Stages of Change Questionnaire (Appendix A) was developed for distribution to

individuals who consented to participate in the study. As shown in Appendix A, we modified the
items on the original questionnaire (i.e., McConnaughy et al., 1983) to make it relevant to the expe-
riences of people who stutter. In most cases, this was accomplished by making minor changes in
wording; for example, the specific term “stuttering” was used in place of “problem,” and “speech
therapy” was used in place of “here.” Similar to the original Stages of Change Questionnaire, this
modified version was a 32-item survey consisting of four, 8-item subscales to measure the four
distinct stages of change: precontemplation, contemplation, action, and maintenance. Participants
were instructed to respond to each item using a 5-point Likert format, where 1 corresponded to
“strongly agree” and 5 corresponded to “strongly disagree”. In accordance with McConnaughy
et al. (1983), the items or statements on the modified questionnaire proposed to assess precon-
templation were: 1, 5, 11, 13, 23, 26, 29, 31; items proposed to assess contemplation were: 2, 4,
8, 12, 15, 19, 21, 24; items proposed to assess action were: 3, 7, 10, 14, 17, 20, 25, 30; and items
proposed to assess maintenance were: 6, 9, 16, 18, 22, 27, 28, 32.
The psychometric properties of the Stages of Change Questionnaire have been evaluated in
a number of studies across a wide variety of substance-related disorders (e.g., Pantalon, Nick,
Franforter, & Carroll, 2002). In particular, it has been shown to have good concurrent and predictive
validity (e.g., Blanchard, Morgenstern, Morgan, Labouvie & Bux, 2003; DiClemente & Hughes,
1990) and internal consistency ranging from acceptable to good (Carey, Purnine, Maisto & Carey,
1999; Pantalon et al., 2002).

5.3. Data analyses

The main purpose of this study was to assess the validity of a stages model to characterize
the change process in the population of adults who stutter, and specifically, to examine the extent
to which responses to items on a modified version of the Stages of Change Questionnaire (or
URICA) would yield a factor structure similar to that which has been observed for other clinical
populations. In order to allow for such an examination, both a confirmatory and exploratory
analyses were used.
The factors in a confirmatory factor analysis are defined as a combination of related or inter-
dependent variables. These underlying factors are identified through cross-correlations among
104 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

the items related to a given factor. For the present study, the modified questionnaire structure
(e.g., McConnaughy et al., 1983) was used as a template; that is, the confirmatory factor analysis
was completed to assess the extent to which the 32 items on the modified questionnaire grouped
together to form the four stages (factors) as observed in previous research (e.g., for the stuttering
group, do questionnaire items 1, 5, 11, 13, 23, 26, and 31 form the precontemplation stage?).
In addition, the confirmatory analysis allowed us to examine relationships between stages; that
is, are adjacent stages more highly correlated than non-adjacent stages? The confirmatory factor
analysis (using EQS v6.1 structural equation modeling software) was completed on the responses
obtained from the questionnaires completed by study participants. Maximum likelihood estima-
tion with robust standard errors was used to adjust for the possibility of incorrect distributional
assumptions (e.g., multivariate non-normality; Bentler, 1995).
The data were also analyzed using an exploratory factor analysis (maximum likelihood esti-
mation with varimax rotation). In contrast to a confirmatory analysis which uses factors (the
four stages in this case) having a predefined relationship with the questionnaire items, an
exploratory factor analysis condenses the variance shared among the factors or variables, and
defines the number of factors to be used based solely on mathematical criteria. For this study,
we used an exploratory factor analysis to delineate the smallest number of factors that could
account for the most variance in the data; therefore, the analysis was not constrained by the
four stages of change, and the manner in which the variables (the questionnaire items) were
combined to form factors was also not controlled by the investigators. We were interested in
finding the best-fitting factor structure for the responses we obtained on the questionnaire (is
the four factor or stage model the “best fit” for our data, or for the population of people who
stutter?).
It should be noted that while an oblique rotation method (oblimin) was initially used in the
maximum likelihood analysis, we decided to report the results from an identical analysis using an
orthogonal, or varimax, rotation. This was done so that we could maximize the clarity with which
the strongest indicators of each factor could be identified. The chi square value obtained from
the confirmatory analysis was compared to the chi square value obtained from the exploratory
(maximum likelihood) analysis to determine the questionnaire structure that best fit the data
gathered in this study. Eigen values were compared to Eigen values from a parallel analysis (Horn,
1965; Zwick & Velicer, 1986) to determine the number of factors to be used in the exploratory
analyses. In addition, a comparison of goodness-of-fit chi square values from vested factor models
(i.e., 3- and 4-factor models) was performed to help determine the appropriate number of factors
to include.

6. Results

6.1. Confirmatory factor analysis

Table 2 contains the factor loading matrix generated by the confirmatory analysis. As shown
in Table 2, 26 of the 32 items on the modified questionnaire were significantly related to their
hypothesized stages (z > 1.96). While the confirmatory analysis verified that the interpretive ques-
tionnaire structure we used fit the data, a goodness of fit test indicated that it was not a perfect fit
(x2 = 1213.233, d.f. = 458, p < .001; comparative fit index = .872; Satorra-Bentler scaled x2 = 572;
d.f. = 458).
Confirmatory analysis results were also used to assess relationships between stages. According
to the stages of change model, stages are additive. This implies that stages adjacent to each
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 105

Table 2
Confirmatory analysis factor loading matrix (correlations between questionnaire items and stages)
Questionnaire item # Stage of change

Precontemplation Contemplation Action Maintenance

1 .57
2 .4
3 .49
4 .34
5 .73
6 .84
7 .70
8 .54
9 .63
10 .27
11 .32
12 .86
13 .42
14 .78
15 .69
16 .41
17 .34
18 .27a
19 .43
20 .64
21 .74
22 .71
23 −.14a
24 .78
25 .70
26 .34
27 .13a
28 .52
29 .20a
30 .67
31 .21a
32 .10a

Note: Correlations not reported were fixed to 0.

a r values not significant (z < 1.96).

other should be more highly correlated than non-adjacent stages. This pattern was observed for
contemplation and action, and for action and maintenance (Table 3). Exceptions to this can be seen
in the unexpected negative correlation between precontemplation (stage 1) and contemplation,
and positive correlation between contemplation and maintenance.

6.2. Exploratory factor analysis

Table 4 shows the maximum likelihood factor loading matrix obtained from the exploratory
analysis, containing the 32 questionnaire items and each of the four factors. Briefly, an exploratory
analysis using maximum likelihood estimation was performed on the 32 × 32 (number of
questionnaire items) correlation matrix to assess the extent to which an alternative interpre-
tive factor structure would provide a better fit to the data (e.g., the 32 items would remain
106 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

Table 3
Correlations between stages
P C A M

P −1.00a −.631 −.954

C .628 1.00a
A .531

P: precontemplation; C: contemplation; A: action; M: maintenance.

a Correlations set at boundaries by software during iterative model fit process. Statistical significance cannot be

estimated.

Table 4
Exploratory analysis factor loading matrix (correlations between questionnaire items and factors)
Questionnaire item # Factors

1 2 3 4

1 −.33 −.18 −.11 −.66

2 .42 .09 −.04 .14
3 .13 .20 .45 .38
4 .09 .65 .17 .20
5 −.83 −.32 .12 .23
6 .16 .81 .06 .50
7 .70 .01 .24 .21
8 .83 .08 .28 .05
9 .23 .52 −.12 .17
10 .22 −.09 .54 .43
11 −.18 −.61 −.19 .05
12 .71 .59 −.08 .05
13 −.07 −.52 −.15 −.16
14 .66 .27 .56 −.19
15 .70 .27 .21 .02
16 .09 .37 −.15 .19
17 .09 .12 .51 −.01
18 .09 .11 .06 .34
19 .07 .26 −.22 .73
20 .74 .31 .02 .11
21 .83 .35 −.22 .07
22 .33 .56 −.15 .43
23 .05 .16 −.54 −.02
24 .61 .65 −.08 .14
25 .27 .63 .48 −.46
26 −.24 −.32 −.22 −.37
27 −.17 .39 .39 .39
28 .24 .52 −.08 .06
29 −.20 .08 −.08 −.03
30 .27 .04 .82 −.08
31 −.34 −.06 −.08 −.10
32 −.23 .43 .36 .09
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 107

the same, but they would be grouped into different factors/stages). A comparison of the chi
square value obtained in the confirmatory analysis (x2 = 1213.233, d.f. = 458), with the chi square
value obtained in the exploratory analysis (x2 = 540.776, d.f. = 374) revealed this to be the case
2 = 673 = 673; p < .01); that is, an interpretive questionnaire structure derived from the
(x84
exploratory analysis provided a significantly better fit to the observed data than the original
interpretive questionnaire structure. A subsequent comparison of Eigen values from an initial
maximum likelihood analysis (no stage or factor structure applied) to criterion Eigen values from
a parallel analysis (Horn, 1965; Zwick & Velicer, 1986), demonstrated that either a 3-factor
or a 4-factor model would be justified for the observed data; however, chi-square goodness
of fit statistics demonstrated that a 4-factor model provided a significantly better fit for the
data than a 3-factor model (difference in x2 = 63.572, d.f. = 29; p = .002). The 4-factor model
obtained from the exploratory analysis represented an increase over the total amount of vari-
ance accounted for when using the original structure in the confirmatory analysis (58% versus
23.2%).

7. Discussion

The principal finding of the present study was that the item groupings in the modified Stages
of Change Questionnaire provided a relatively good fit for the participant responses. Of the 32
items on the questionnaire, 26 demonstrated a significant relationship with their hypothesized
stage.
A second important finding was the mixed results for the between-stage correlation analysis
(see Table 3). For example, contemplation and action were positively correlated with their adja-
cent stages (i.e., action and maintenance, respectively), an observation that is not surprising if
one assumes the validity of an additive stage model (i.e., individuals tend to move through stages
in sequence). Support for this assumption is also evident in the negative correlation observed
between the nonadjacent stages precontemplation and action; that is, participants who responded
either “strongly agree” or “somewhat agree” to items associated with precontemplation, scored
low on action items (i.e., “strongly disagree” or “somewhat disagree”), and vice versa. This
observation supports the premise that people in the precontemplation stage are likely to deny
a need for change, while those in the action stage are actively working to maintain change.
Alternately, as shown in Table 3, there was also a negative correlation between precontempla-
tion and each of the adjacent three stages (i.e., contemplation, action, and maintenance). As
an adjacent stage, one would expect that contemplation would be positively correlated with
precontemplation, but that was not the case in the present study. As well, one would expect
that maintenance and contemplation would be negatively correlated, but that was also not the
case. These findings suggest, among other things, that a discriminant or cluster analysis might
reveal that individuals who stutter who are “precontemplators” represent a distinct subgroup of
individuals.
Taken together, the results of both the confirmatory and exploratory factor analysis indicate that
a questionnaire incorporating items that better reflect the unique behavioral, cognitive and affective
variables that characterize stuttering may better discriminate stages of change in people who stutter
as they move through therapy, or are engaged in self-directed change. The exploratory factor
analysis yielded a correlation matrix (Table 4) that suggested that this alternative questionnaire
structure may also include four factors. While it is beyond the scope of the present study to design
and assess the validity of an alternative questionnaire, the results of the exploratory analysis can
guide the process of developing an appropriate instrument.
108 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

7.1. Development of an alternative questionnaire structure

A reasonable place to start the process of alternative questionnaire development is to obtain

data from additional subjects in order to validate the current factor loadings from the exploratory
analysis. Next, factor patterns derived from the maximum likelihood correlation matrix presented
in Table 4 should be examined. First, as shown in Table 4, there is a wide range of item loadings
for each of the four factors (e.g., .050 (item 23) to .83 (item 8) for factor #1; .01 (item 7) to .81
(item 6) for factor #2, and so on). Second, if we use r > .40 as a conservative cutoff value, we
see that of the 32 questionnaire items, 22 were correlated with a single factor. The remaining
ten items were either correlated with more than one factor (e.g., item 27), or were not correlated
with any of the four factors (e.g., item 18). The reason that some items were not loaded on
a single factor is unknown, but a likely explanation is that there is overlap between factors.
Alternately, items that did not correlate with any of the factors suggest that there may be additional
factors that are not adequately represented by the existing questionnaire items. It may also be
the case that these items are not representative of any salient factor and should be eliminated.
Given the relatively small sample size, however, it does not seem reasonable to eliminate any
items without additional data. Although the general structure of the factor analytic solutions
presented here can be considered reasonably stable (e.g., MacCallum, Widaman, Zhang, & Hong,
1999), the confidence intervals surrounding factor loadings are quite broad and would likely
lead one to discard good indicators of a factor and/or include poor indicators of a factor due to
chance.
The items correlated with each factor should also be examined as a group, to determine the
extent to which they suggest specific themes that may characterize different but related ‘stages’
within a process of change related to stuttering. Tables 5–8 present the item-factor groupings
as determined by the results of the exploratory factor analysis presented in Table 4. Of note
is that the items in bold were also loaded on similar or “mirror” factors (i.e. stages) in the
confirmatory analysis. For example, item 1 was loaded on Factor 1 in the exploratory analysis,
and precontemplation in the confirmatory analysis.

Table 5
Exploratory factor analysis: questionnaire items grouped in Factor 1
Item #

8 I’ve been thinking that I might want to change something about my speech
21 Maybe speech therapy will be able to help me
20 I have started working on my stuttering but I would like help
12 I’m hoping speech therapy will help me to better understand my stuttering
7 I am finally doing some work on my stuttering
15 I have a problem and I really think I should work on it
14 I am really working hard to change
24 I hope that speech therapy will have some good advice for me
2 I think I might be ready to improve my speech
22 I may need a boost right now to help me maintain the changes I’ve already made
1a As far as I’m concerned, my speech does not need changingb
31a I would rather cope with my stuttering than try to change itb
5a I don’t have a problem with stuttering, It doesn’t make sense for me to be in therapyb
a Items in bold were grouped in precontemplation stage in confirmatory analysis.
b Items are inversely correlated to the factor.
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 109

Table 6
Exploratory factor analysis: questionnaire items grouped in Factor 2
Item #

6 It worries me that I might slip back on stuttering I have already changed so I would like to seek help
4a It might be worthwhile to work on my stuttering
24a I hope that speech therapy will have some good advice for me
25 Anyone can talk about changing; I’m actually doing something about it
12 I’m hoping speech therapy will help me to better understand my stuttering
22a I may need a boost right now to help me maintain the changes I’ve already made
9 I have been successful in working on my stuttering but I’m not sure I can keep up the effort on my own
28 It is frustrating, but I feel I might be having a recurrence of the stuttering I thought I had resolved
32 After all I had done to try to change my stuttering, every now and again it comes back to haunt me
27 I would like to prevent myself from having a relapse of stuttering
16 I’m not following through with what I had already changed as well as I had hoped, and I would like to
prevent a relapse of my stuttering
21 Maybe speech therapy will be able to help me
20 I have started working on my stuttering but I would like help
26 R: All this talk about stuttering is boring, why can’t people just forget about their stuttering?
5 R: I don’t have a problem with stuttering, It doesn’t make sense for me to be in therapy
13 I guess I have faults, but there’s nothing that I really need to changeb
11 Being in therapy is pretty much a waste of time for me because stuttering doesn’t have to do with meb
a Items in bold were grouped in contemplation stage in confirmatory analysis.
b Items are inversely correlated to the factor.

As shown in Table 5, the first factor (Factor 1) was associated with 13 of the 32 questionnaire
items, with three of these items (1, 31, and 5) correlated in the opposite direction as all the other
items in the factor grouping. With the exception of these three items, a cursory examination of the
items grouped under Factor 1 suggests a general theme of therapy preparedness or consideration
of the potential benefits of therapy for stuttering. As shown in Table 6, the items grouped under
Factor 2 vacillate between two contrasting themes. For example, items 4 and 25 (“It might be
worthwhile to work on my stuttering”, and “Anyone can talk about changing; I’m actually doing
something about it.”) suggest themes of self-efficacy and optimism, while items 9 and 28 (“I have
been successful in working on my stuttering but I’m not sure I can keep the effort up on my own”,

Table 7
Exploratory factor analysis; questionnaire items grouped in Factor 3
Item #

30a I am actively working on my stuttering

14 I am really working hard to change
10a At times my stuttering is difficult, but I’m working on it
17a Even though I’m not always successful in changing, I am at least working on my stuttering
25a Anyone can talk about changing; I’m actually doing something about it
3a I am doing something about my stuttering that has been bothering me
27 I would like to prevent myself from having a relapse of stuttering
32 After all I had done to try to change my stuttering, every now and again it comes back to haunt me
23 I may be part of the problem, but I really don’t think I amb
a Items in bold were grouped in action stage in confirmatory analysis.
b Items are inversely correlated to the factor.
110 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

Table 8
Exploratory factor analysis: questionnaire items grouped in Factor 4
Item #

19 I wish I had more ideas on how to solve my stuttering

6a It worries me that I might slip back on stuttering I have already changed so I would like to seek help
10 At times my stuttering is difficult, but I’m working on it
22 I may need a boost right now to help me maintain the changes I’ve already made
27a I would like to prevent myself from having a relapse of stuttering
3 I am doing something about my stuttering that has been bothering me
18a I thought once I had resolved the stuttering, I would be free of it, but sometimes I still find myself
struggling with it
26 All this talk about stuttering is boring, why can’t people just forget about their stuttering?b
25 Anyone can talk about changing; I’m actually doing something about itb
1 As far as I’m concerned, my speech does not need changingb
a Items in bold were grouped in maintenance stage in confirmatory analysis.
b Items are inversely correlated to the factor.

and “It is frustrating, but I feel I might be having a recurrence of the stuttering I thought I had
resolved.”) imply fear of failure. Finally, examination of the item-factor relationships displayed
in Tables 7 and 8 suggest themes of “responsibility” or “effort to change” in Factor 3, and “need
for additional options or resources” for Factor 4.
In the present study, we made relatively small changes in the Stages of Change Questionnaire
so as to make the statements relevant to stuttering and the attitudes and beliefs that people who
stutter might experience. While this slightly modified questionnaire fit the data, analysis of the
item-factor groupings from the exploratory factor analysis (Tables 5–8) indicate that the fit would
be improved if the instrument contained additional items that relate to the unique aspects of the
problem of stuttering. For example, Turnbull (2000) suggested the addition of statements to the
original Stages of Change Questionnaire that reflect feelings of being “fed up” with stuttering
(e.g., “I feel as though I’ve had enough with regard to my stuttering.”), as a valid way to indicate
movement from one stage to the next. The expressed feeling of being “fed up” might suggest that
the individual is moving away from a general preparedness or consideration of therapy benefits
(i.e., Factor 1) into self-efficacy and optimism (i.e., Factor 2) and beyond. In addition, statements
that allow individuals to indicate the extent to which they feel favorably toward the clinician and
the therapy approach would likely add significantly to the instrument and the analysis. Further, it
is possible that there are items that could be added to an alternative questionnaire for the parents
of young children who stutter so as to obtain an indication of where they are in the change
process. This appears especially important given that most therapy approaches for preschool
children who stutter involve a considerable amount of parental involvement. The importance of
questionnaire items, or statements, that are more relevant to the stuttering population obviously
remains to be seen; however, additional modifications to the preliminary questionnaire used here
seems necessary in future studies of the applicability of a stages of change model to stuttering
therapy.

7.2. Matching therapy strategies to “stages” in stuttering therapy

A significant problem in planning stuttering treatment is deciding when to introduce specific

techniques, or address the different components of stuttering, so as to maximally facilitate change
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 111

and a positive outcome. Results from the present study provide support for Turnbull’s (2000) sug-
gestion that the stages of change model can provide guidelines for delineating specific strategies
or techniques that would help individuals move successfully through different stages of stuttering
therapy. While the treatment strategies or techniques conventionally used in stuttering therapy
differ from the processes used in psychotherapy and other treatment approaches used for addic-
tive or socially unacceptable behaviors, they appear similar to most of the processes first defined
by Prochaska and DiClemente (1986). Similar to the processes shown in Table 1, well-established
techniques used in stuttering intervention differentially address experiences (thoughts and emo-
tions), and behaviors (speech; e.g., Healey, Trautman & Susca, 2004; Shapiro, 1999). For example,
strategies that address thoughts and feelings include examining attitudes about talking and stut-
tering (e.g., assessing how one feels and thinks about oneself with respect to a problem: value
clarification, corrective emotional experience, or precontemplation; see Table 1), desensitization,
and self-disclosure.
Perz, DiClemente and Carbonari (1996), describing the processes related to successful cessa-
tion from smoking, discovered that it was more appropriate to use experiential processes during
particular stages of change, and behavioral processes during others. Experiential processes address
affective and cognitive components, whereas behavioral processes are ones that teach the specific
behaviors necessary for change to come about. Examples of experiential processes are conscious-
ness raising, self re-evaluation, and self and social liberation (see Table 1). Perz et al. (1996)
found that it was more beneficial to treatment progress if clients in the precontemplation and con-
templation stages used the experiential processes of consciousness-raising and self re-evaluation.
On the other hand, individuals in the action stage benefited more from such behavioral processes
as counter conditioning and stimulus control. Using processes at the “right” time were shown to
facilitate movement from one stage to another.
The work of Perz et al. (1996) and Turnbull (2000) provide direction for matching stut-
tering therapy techniques or tools to “where” a person who stutters might be in the change
process. For example, if it is determined that a person who stutters is in the “thinking about
therapy” stage, he or she would likely benefit from “bibliotherapy”, or reading information per-
taining to talking and stuttering. This is an experiential activity that serves as a mechanism
for raising consciousness. Such experiential processes could also be used by those who are
preparing for therapy, or those who are considering the potential benefits of therapy. These indi-
viduals are likely to benefit from activities such as exploring the implications of change, and
weighing the advantages and disadvantages of both staying the same and of changing (self re-
evaluation).
For those who are making an active effort to change, Turnbull again borrows from the pro-
cesses believed to support movement through action to maintenance. For example, individuals
who stutter, who are actively involved in the change process would be primed, or ready to engage
in the use of techniques or strategies for changing speech behaviors (i.e., strategies specific to
either fluency shaping or stuttering modification; “substituting alternatives for problem behaviors”
(see Table 1). In addition, in this “action” stage, strategies to facilitate thoughts and feelings that
promote change would be important, and include, among other things, relaxation and assertive-
ness training, desensitization work, and enlisting others to provide support and encouragement,
such as one receives from support groups or self-help organizations. These are all commonly
used strategies in therapy for stuttering. Looking at self-help organizations from a different per-
spective, however, it might be the case that participation in a support group, while intuitively
helpful, has the opposite effect on those people who believe that there is no problem, or that
therapy or change is not presently needed. For these individuals, attending a support group may
112 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

be a confrontational or intimidating experience in that they do not share (or see) common expe-
riences with other group members. In these cases, the individual likely has nothing to share
and may come away frustrated and “turned-off” to stuttering intervention of any kind. Finally,
Turnbull suggests that individuals in the “maintenance” stage of change, or who are looking
for additional options for change that extend beyond therapy, need to develop a “tool box” of
strategies to help maintain change. Using the processes associated with the “maintenance” stage
of change (see Table 1), this “tool box” might include the expansion of support networks, self-
advocacy, rewarding oneself for making changes, and consideration of relapse with an open
mind.

7.3. Limitations of the present study

As previously discussed, the most obvious limitation of the present study is the relatively
small number of participants, especially given the nature of factor analysis. Undoubtedly future
research should attempt to collect additional data from a larger number of subjects. While some
statisticians suggest a minimum of 5–15 participants per variable to increase the reliability of
factor analyses, others have disagreed that this is necessary. For example, MacCallum et al.
(1999) and others have argued that subject to item ratio is not necessarily a valid indicator of
the reliability of an individual factor analytic solution. In their Monte Carlo study, MacCallum et
al. observed that the congruence of a factor analytic solution from a sample to a population was
a complex function of sample size, communalities (i.e., the sum of squared factor loadings for
a variable), and factor overdetermination (i.e., the number of items serving as indicators of the
factor). Among these three, communalities were the most important. Using this observation, the
results of the present study have wide communalities (mean: .52; 5th to 95th %ile; range: .13–.88),
and high factor overdetermination (with eight items per factor, compared with 6.66 items per
factor in McCallum et al.’s study). With the present sample size of 44, we expect that the results
of the exploratory factor analysis would therefore have high congruence (>.8) with population
values.
Besides sample size, another potential limitation of this study is the population from which
participants were drawn, and the extent to which this may have influenced the results of
the analysis, particularly the confirmatory factor analysis. Recall that the confirmatory anal-
ysis yielded relatively strong correlations between the maintenance stage and the other three
stages (precontemplation, contemplation and action). This correlation pattern was unexpected,
as typically, nonadjacent stages (e.g., maintenance and precontemplation, or maintenance and
contemplation) are not correlated. In the present study, subjects were recruited from either
speech-language pathologists or self-help organizations (i.e., NSA), and all reported a history
of stuttering therapy (as might be expected from their association with so-called gatekeepers).
For this reason it is safe to say that a disproportionate number of subjects may be represen-
tative of either action or maintenance stages of change, and such unequal distribution across
stages may make it unreasonable to expect sphericity (i.e., higher correlations between adja-
cent stages) in the confirmatory factor analysis. We cannot know if this is the case based
on one data set, but this possibility can be ruled out in future research by comparing the
factor inter-correlation matrices from separate confirmatory factor analyses, one of which is
obtained from subjects who are “therapy graduates” and one from subjects who have never
had formal stuttering therapy (or have not been enrolled in therapy for an extended period of
time).
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 113

Appendix A. Modified “Stages of Change” Questionnaire (after McConnaughy et al.,

1983)
114 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120
J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 115
116 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120
J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 117
118 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

CONTINUING EDUCATION

Stages of change and stuttering: A preliminary view

QUESTIONS

1. The underlying premise of the “stages of change” or “transtheoretical model” is

a. that there are various components to stuttering treatment
b. that there are commonalities across treatment approaches in the processes people use to
change their behavior
c. that change is inevitable
d. that some treatment approaches are better than others
e. that change processes are only seen in nonaddictive behaviors
2. The “stages of change” can be seen in
a. individuals enrolled in treatment programs
b. so-called “self-changers”
c. both individuals enrolled in therapy, and self-changers
d. primarily females
e. primarily males
3. In general, movement through the “stages of change” is
a. linear
b. nonadjacent
c. uncorrelated
d. spiral
e. random
4. Results from the present study suggest that
a. the stages of change are not valid
b. adjacent stages of change are not correlated
c. change can be predicted by stuttering severity
d. measurements of stuttering are not reliable
e. a modified Stages of Change Questionnaire structure provides a relatively good fit to the
responses obtained from people who stutter
5. “Self-liberation” is a
a. level of change
b. confounding element to change
c. degree of change
d. way to describe people who stutter who make progress in therapy
e. process of change

References

Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckman (Eds.), Action control:
From cognition to behaviour (pp. 11–39). New York: Springer-Verlag.
Armitage, C. J. (2006). Evidence that implementation intentions promote transitions between the stages of change. Journal
of Consulting and Clinical Psychology, 74, 141–151.
Bentler, P. M. (1995). EQS Structural Equations Program Manual. Encino, CA: Multivariate Software, Inc.
 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120 119

Blanchard, K., Morgenstern, J., Morgan, T., Labouvie, E., & Bux, D. (2003). Motivational subtypes and continu-
ous measures of readiness for change: Concurrent and predictive validity. Psychology of Addictive Behaviors, 17,
56–65.
Blood, G. W. (1993). Treatment efficacy in adults who stutter: Review and recommendations. Journal of Fluency Disorders,
18, 303–318.
Brownell, K., Marlatt, G., Lichtenstein, E., & Wilson, T. (1986). Understanding and preventing relapse. American
Psychologist, 41(7), 765–782.
Carey, K., Purnine, D., Maisto, S., & Carey, M. (1999). Assessing readiness to change substance abuse: A critical review
of instruments. Clinical Psychology: Science and Practice, 6, 245–266.
Craig, A. (1998). Relapse following treatment for stuttering: A critical review and correlative data. Journal of Fluency
Disorders, 23, 1–30.
DiClemente, C. C. (1993). Changing addictive behaviors: A process perspective. Current directions in Psychological
Science, 29(4), 101–106.
DiClemente, C., & Hughes, S. (1990). Stages of change profiles in outpatient alcoholism treatment. Journal of Substance
Abuse, 2, 217–235.
Guitar, B. (1976). Pretreatment factors associated with the outcome of stuttering therapy. Journal of speech and hearing
Research, 19, 590–600.
Guitar, B. (1998). Stuttering: An integrated approach to its nature and treatment (2nd ed.). Baltimore: Williams &
Williams.
Guitar, B., & Bass, C. (1978). Stuttering therapy: The relation between attitude change and long-term outcome. Journal
of Speech and Hearing Disorders, 43, 392–400.
Healey, E. C., Trautman, L. S., & Susca, M. (2004). Clinical applications of a multidimensional approach for the assessment
and treatment of stuttering. Contemporary Issues in Communication Science and Disorders, 31, 40–48.
Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–185.
Huinck, W. J., Langevin, M., Kully, D., Graamans, K., Peters, H. F. M., & Hulstijn, W. (2006). The relationship between
pre-treatment clinical profile and treatment outcome in an integrated stuttering program. Journal of Fluency Disorders,
31(1), 43–63.
MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample size in factor analysis. Psychological Methods,
4, 84–99.
Manning, W. H. (2001). Clinical decision making in fluency disorders (2nd ed.). Vancouver, Canada: Singular.
Manning, W. H. (2006). Therapeutic change and the nature of our evidence: Improving our ability to help. In N. Bernstein
Ratner & J. A. Tetnowski (Eds.), Current issues in stuttering research and practice. Mahwah, NJ: Lawrence Erlbaum
Assoc.
Marcus, B. H., & Simkin, L. R. (1994). The transtheoretical model: Applications to exercise behavior. Medicine and
Science in Sports and Exercise, 26(11), 1400–1404.
McConnaughy, E. A., Prochaska, J. O., & Velicer, W. F. (1983). Stages of change in psychotherapy: Measurement and
sample profiles. Psychotherapy: Theory, Research, and Practice, 20(3), 368–375.
Pallonen, E., Leskinen, L., Prochaska, J. O., Willey, J., Kaariainen, R., & Salonen, J. T. (1994). Self-help smoking cessa-
tion manual intervention among middle-aged Finnish men: An application of the transtheoretical model. Preventive
Medicine, 23(4), 507–514.
Pantalon, M., Nich, C., Frankforter, T., & Carroll, K. (2002). The URICA as a measure of motivation to change among
treatment-seeking individuals with concurrent alcohol and cocaine problems. Psychology of Addictive Behaviors,
16(4), 299–307.
Pantalon, M., & Swanson, A. (2003). Use of the Rhode Island Change Assessment to measure motivational readiness to
change in psychiatric and dually diagnosed individuals. Psychology of Addictive Behaviors, 17(2), 91–97.
Perz, C. A., DiClemente, C. C., & Carbonari, J. P. (1996). Doing the right thing at the right time? The interaction of stages
and processes of change in successful smoking cessation. Health Psychology, 15(6), 462–468.
Prochaska, J. O. (1999). How do people change and how can we change to help many more people? In M. A. Hubble, B.
L. Duncan, & S. D. Miller (Eds.), The heart & soul of change: What works in therapy (pp. 227–255). Washington,
D.C.: American Psychological Association.
Prochaska, J. O., & DiClemente, C. C. (1984). The transtheoretical approach: Crossing traditional boundaries of therapy.
Homewood, IL: Dow Jones-Irwin.
Prochaska, J. O., & DiClemente, C. C. (1986). The transtheoretical approach. In J. C. Norcross (Ed.), Handbook of eclectic
psychotherapy. New York: Brunner/Mazel.
Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how people change: Applications to addictive
behaviors. American Psychologist, 47(9), 1102–1114.
120 J. Floyd et al. / Journal of Fluency Disorders 32 (2007) 95–120

Prochaska, J. O., Velicar, W. F., Rossi, J. S., Goldstein, M. G., Marcus, B. H., Rakowski, W., et al. (1994). Stages of
change and decisional balance for 12 problem behaviors. Health Psychology, 13(1), 39–46.
Ruggiero, L. (1998). Transtheoretical model: Applications in the prevention and treatment of cancer. Medical and Pediatric
Oncology supplement, 1(69), 74.
Sarafino, E. (2006). Health Psychology:Biopsychosocial Interactions. Hoboken, NJ: Wiley & Sons, Inc.
Shapiro, D. A. (1999). Stuttering intervention: A collaborative journey to fluency freedom. Austin: TX: Pro-Ed.
Suris, A. M., Carmen Trapp, M., DiClemente, C. C., & Cousins, J. (1998). Application of the transtheoretical model of
behavior change for obesity in Mexican-American women. Addictive Behaviors, 23(5), 655–668.
Turnbull, J. (2000). The transtheoretical model of change: Examples from stuttering. Counseling Psychology Quarterly,
31(1), 13–21.
Zebrowski, P. M., & Conture, E. G. (1998). Influence of nontreatment variables on treatment effectiveness for school-age
children who stutter. In A. K. Cordes & R. J. Ingham (Eds.), Treatment efficacy for stuttering: A search for empirical
bases (pp. 293–310). San Diego, CA: Singular Publishing Group, Inc.
Zwick, W. R., & Velicer, W. F. (1986). Comparison of five rules for determining the number of components to retain.
Psychological Bulletin, 99(3), 432–442.
Jennifer Floyd graduated from the University of Colorado at Boulder with majors in Psychology and Biology. She
attained her MA in speech-language pathology from the University of Iowa in 2001. Since then, Jennifer has worked in
the elementary school system in Illinois, and is currently in private practice in Colorado.

Patricia M. Zebrowski is an associate professor in Speech Pathology and Audiology at the University of Iowa. She is
a Fellow of the American Speech-Language and Hearing Association, and a Board Recognized Fluency Specialist. Her
interests are in the area of stuttering, with specific emphasis on children who stutter.

Gregory A. Flamme is an assistant professor in Speech Pathology and Audiology at Western Michigan University. He
earned a PhD in Audiology (University of Memphis) and completed postdoctoral studies in Epidemiology and Biostatistics
(University of Iowa). His research focuses on hearing loss prevention and treatment.