Week 8 Assignment

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=wswp20

Journal of Social Work Practice in the Addictions

ISSN: 1533-256X (Print) 1533-2578 (Online) Journal homepage: https://www.tandfonline.com/loi/wswp20

Theories of Motivation in Addiction Treatment: Testing the Relationship of the Transtheoretical Model of Change and Self-Determination Theory

Kerry Kennedy PhD & Thomas K. Gregoire PhD

To cite this article: Kerry Kennedy PhD & Thomas K. Gregoire PhD (2009) Theories of Motivation in Addiction Treatment: Testing the Relationship of the Transtheoretical Model of Change and Self-Determination Theory, Journal of Social Work Practice in the Addictions, 9:2, 163-183, DOI: 10.1080/15332560902852052

To link to this article: https://doi.org/10.1080/15332560902852052

Published online: 13 May 2009.

Submit your article to this journal

Article views: 3537

Citing articles: 7 View citing articles

163

Journal of Social Work Practice in the Addictions, 9:163–183, 2009 Copyright © Taylor & Francis Group, LLC ISSN: 1533-256X print/1533-2578 online DOI: 10.1080/15332560902852052

WSWP1533-256X1533-2578Journal of Social Work Practice in the Addictions, Vol. 9, No. 2, March 2009: pp. 1–38Journal of Social Work Practice in the Addictions

Theories of Motivation in Addiction Treatment: Testing the Relationship

of the Transtheoretical Model of Change and Self-Determination Theory

Theories of Motivation in Addiction TreatmentK. Kennedy and T. K. Gregoire

KERRY KENNEDY, PHD Assistant Professor, Department of Social Work and Gerontology, Weber State University,

Ogden, Utah, USA

THOMAS K. GREGOIRE, PHD Associate Professor, College of Social Work, Ohio State University, Columbus, Ohio, USA

This study explored the relationship between 2 theories of motivation: self-determination theory (SDT) and the transtheoretical model of change (TTM), and sought to determine whether the source of moti- vation described by SDT would predict TTM’s stage of change. SDT was operationalized as the level of internal or external motivation for treatment, and TTM was operationalized as 3 stages of change: precontemplation, contemplation, and action. Our data came from the Drug Abuse Treatment Outcome Study published in 2004. A multinomial logistic regression analysis indicated that there was a significant relationship between source of motivation and stage of change at intake. Controlling for severity, treatment history, legal status, and primary substance use, persons entering treatment with higher levels of internal motivation were more likely to be in the action stage than the precontemplation stage. Higher levels of internal motivation also predicted a greater likelihood of being in the contemplation rather than the precontemplation stage.

KEYWORDS addiction, motivation, stages of change

Received July 18, 2006; accepted February 22, 2007. Address correspondence to Kerry Kennedy, Department of Social Work and Gerontology,

Weber State University, 1211 University Circle #148, Ogden, UT 84041, USA. E-mail: kerrykennedy@ weber.edu

164 K. Kennedy and T. K. Gregoire

Recent years have seen an evolution in thought about the role of motivation in substance abuse treatment. Motivation was once viewed as a function of individual differences and largely related to personality traits. Individuals who did not comply with treatment were considered to be unmotivated (Clancy, 1961). Motivation is now considered more a function of an interac- tion of individual and environmental factors (Miller & Rollnick, 2002). Current views of motivation in treatment place the onus for client motivation on the clinician, recognizing that the interaction of the clinician and client “has a crucial impact on how they respond and whether treatment is successful” (Center for Substance Abuse Treatment, 1999, p. 3). Clients once viewed as not ready to change their behavior are now more likely seen as in need of different interventions than those perceived as more motivated (Prochaska, DiClemente, & Norcross, 1992). Interventions such as motivational inter- viewing are based on the demonstrated assumption that the clinician need not wait for a client’s readiness but instead can impact a client’s level of motivation and guide them toward prosocial behavior (Miller & Rollnick, 2002).

The change in the perception of the role of motivation has been heavily influenced by the emergence of the transtheoretical model (TTM; DiClemente, 2003; Prochaska, 1979; Prochaska & DiClemente, 1983; Prochaska et al., 1992). As described later, this theoretical perspective on a client’s readiness to change has led practitioners to recognize that motivation to change might be a malleable, dynamic, and nonlinear process.

Although less well known than TTM, self-determination theory (SDT; Deci & Ryan, 1985) describes the role of internal and external factors in understanding motivation. The two theories provide distinct approaches to understanding the role of motivation in affecting change among persons who misuse substances. This article explores the relationship of these two theories and considers if the two approaches in tandem provide a greater understanding of motivation among persons with substance use disorders.

THE TRANSTHEORETICAL MODEL

The TTM was derived from a compilation of 18 different psychological and behavioral theories and provides a temporal framework for describing intentional behavior change (Prochaska, 1979). Developed originally as a model for understanding client-initiated attempts to modify their nicotine addiction (Prochaska & DiClemente, 1983), TTM has been adapted to a wide range of behaviors including substance use (DiClemente & Hughes, 1990).

DiClemente (2003) described TTM as consisting of four interrelated dimensions of change that include stages, processes, markers, and a context of change. The stages of change are delineated with a time frame and tasks

Theories of Motivation in Addiction Treatment 165

associated with movement through that stage. There are five stages of change (Prochaska et al., 1992). Persons identified in the precontemplation stage have no current intention to modify their behavior, and might not acknowl- edge that they have a problem. Persons characterized as contemplators are thinking about addressing their problem, but have yet to take action. The preparation stage reflects a client’s intention to make some change attempt during the next month. Persons who have made an unsuccessful change effort in the preceding year are also in this stage of change. The action stage is associated with the initiation of modified substance use behavior. Typically, the action stage extends from 3 to 6 months (DiClemente, 2003). During the final maintenance stage, beginning at 6 months after behavior change, persons take steps to avoid relapse and consolidate their new lifestyle.

These stages are accompanied by processes of change. Each process consists of interventions appropriate to assist a person in moving to the next stage of change (Prochaska et al., 1992). The context of change attempts to address other areas that might contribute to maintenance of the problem including interpersonal, social, and environmental dimensions (DiClemente, 2003).

TTM emphasizes the importance of attempting and then maintaining new behavior in understanding motivation for change. The concept of deci- sional balance refers to the weight of the evidence for and against a certain behavior. In TTM, persons initiate change as the balance tips against the benefits of the addictive behavior. TTM does not appear to distinguish between the importance of internal and external sources of motivation with respect to decisional balance, although early interventions do tend to empha- size external pressure.

The literature on the influence of motivation type on stage of change is limited and divided. O’Hare (1996) found that people referred through the court were more likely to be in the precontemplation stage than peo- ple who were not court referred. However, Gregoire and Burke (2004) reported that persons who were legally coerced to outpatient substance abuse treatment were more likely to be in the action stage of change than precontemplation.

A considerable literature has emerged on TTM, with mixed findings on its efficacy as a tool for understanding addiction and recovery. Knowledge of a client’s stage of change has been demonstrated in some studies as an important tool for predicting treatment outcomes. In particular, studies have associated a client’s stage of change with subsequent alcohol use (Heather, Rollnick, & Bell, 1993), heroin or cocaine use (Henderson, Galen, & Saules, 2004; Prochaska et al., 1994), and treatment dropout (Callaghan et al., 2005; Edens & Willoughby, 2000; Haller, Miles, & Cropsey, 2004; Simpson & Joe, 1993). In general, clients identified as precontemplators fared worse on sub- sequent outcomes than those who presented for treatment in more advanced stages of change.

166 K. Kennedy and T. K. Gregoire

However, other authors reported finding no relationship between stage of change and either posttreatment substance use (Belding, Iguchi, & Lamb, 1997; Burke & Gregoire, 2007) or Addiction Severity Index composite scores (Burke & Gregoire). Another study found the stages of change model unable to predict either treatment attendance or the percentage of days abstinent (Blanchard, Morgenstern, Morgan, Labouvie, & Bux, 2003).

Critics of the TTM argue that this model is arbitrary in its division of the stages of change and the time frame of each stage (Sutton, 2001). Bandura (1997) described the TTM as “arbitrary pseudo-stages” (p. 8) and not a true stage theory. Davidson (1998) and Bandura (1997) suggested that the TTM does not include stages that are distinctly, qualitatively different from each other, noting that one stage can be considered an extension of the previous stage. Weinstein, Rothman, and Sutton (1998) also described the specific time points of the stages as arbitrary.

DiClemente (2003) acknowledged that the TTM stage model does not produce fixed stages with either a determinant order or a single linear path- way. Instead the stages are best understood as a developmental model of recovery in which the resolution of earlier tasks impacts the later stages.

SELF-DETERMINATION THEORY

SDT (Deci & Ryan, 1985) has been applied to many areas, such as medica- tion adherence, weight loss, and test-taking behavior in school-aged children. However, the application of this theory specifically to substance abuse has been limited (Ryan, Plant, & O’Malley, 1995; Zeldman, Ryan, & Fiscella, 2004). Nevertheless, the application of SDT might complement TTM as a framework for understanding motivation in substance misuse treatment. Whereas TTM does not place a major emphasis on the internal or external aspects of a client’s motivation to change, SDT addresses the source of motivation specifically by outlining a framework for understanding internal and external sources of motivation and the impact of each on treatment out- comes. SDT defines motivation as consisting of six categories that describe a continuum of external to internal motivation.

SDT postulates that persons with higher internal motivation should have better treatment outcomes, and that high external motivation in the absence of internal motivation is associated with less positive outcomes. Zeldman et al. (2004) found that clients in a methadone maintenance pro- gram with higher levels of internal motivation had lower relapse rates and higher program participation than externally motivated persons.

Ryan et al. (1995) also found that higher internalized motivation was negatively correlated with dropping out of treatment. Their study also observed an interaction between internal and external motivation, as per- sons with both high internal and high external motivation were most likely

Theories of Motivation in Addiction Treatment 167

to persist in treatment. External motivation was positively related to out- comes only when internal motivation was present.

RELATIONSHIP BETWEEN SDT AND TTM

Both theories make important contributions to understanding motivation among persons with substance use disorders. SDT identifies the forces that might influence an individual to initiate behavior change and the psycho- logical mechanisms that drive an individual to accomplish change. TTM identifies the change processes in individuals within a temporal dimension of motivation and provides a framework for understanding movement toward behavior change.

Together, SDT and the TTM offer a more comprehensive view of moti- vation than either do independently. Whereas SDT emphasizes the influ- ence of perceived antecedents of behavior change on the individual, TTM provides structure to understand how clients move through a progression of behavior change. Although both theories incorporate information about the source of motivation, SDT more explicitly distinguishes the influence of internal and external sources of motivation. TTM refers to two markers of change—decisional balance and self-efficacy—and suggests that early influ- ences to change tend to be external as the decisional balance scale is tipped. TTM suggests that internal forces are at work as self-efficacy is acquired and change is maintained. However, efficacy refers less explicitly to motivation and more to the development of a confidence that one might succeed in a change effort.

Despite the utility of each in informing a greater understanding of motivation, a specific relationship between these two theories has yet to be determined within a population of persons who misuse substances. The purpose of this research is to explore the relationship of TTM and SDT among a sample of substance use treatment participants. In particu- lar, we seek to model the relationship between source of motivation as described by SDT and TTM’s stage of change to test the following hypotheses: (a) Higher levels of external motivation will predict mem- bership in the precontemplation stage of change; and (b) greater internal motivation will predict later stages of change, specifically contemplation and action.

METHOD

Data for this study were obtained from the Drug Abuse Treatment Outcome Study–Adult (DATOS) published in 2004 (U.S. Department of Health and Human Services, National Institute on Drug Abuse, 2004). The DATOS study

168 K. Kennedy and T. K. Gregoire

was conducted by Research Triangle Institute and funded by the National Institute on Drug Abuse. DATOS data were derived from a longitudinal pro- spective cohort design of 10,010 persons aged 18 or older. Clients from 96 programs in 11 cities were purposively chosen and interviewed at intake and again at several points during treatment. The selection of communities and programs reflected typical drug treatment programs in medium and large-sized U.S. cities and consisted of both publicly and privately funded entities.

Data were collected from 1991 to 1993 via face-to-face interviews con- ducted with participants in four different treatment modalities: outpatient methadone maintenance, long-term residential, outpatient drug free, and short-term inpatient. Data for our study consisted of two waves of interviews. The first wave was from an initial interview conducted as soon as possible after admission to treatment and the second wave came from an interview that was conducted about 1 week later.

Participants

Of the 10,010 participants in the total sample, 8,725 had a completed first and second intake interview on record. Although the initial two waves of DATOS data consisted of a purposive sample of treatment admissions, data for the DATOS follow-up interviews were derived from a random sample. The sample employed for our analysis consisted of a random sample drawn from those individuals with completed Wave 1 and 2 data. Table 1 describes the characteristics of the sample in greater detail.

The majority of the 4,347 participants were male (67.1%), and most were over the age of 30 (58.2 %), with a mean age of 32.5 (SD = 7.4). African Americans made up 46.7% of our sample, 38.7% were White, and 11.9% were Hispanic. Most respondents (55.0%), had no current legal involvement, although 31.9% reported being on probation or parole.

Most of the respondents reported being unmarried at the time of their first interview; 45.4% were never married, 19.5% were currently married, 12.1% were living as married, and 13.5% reported being divorced. In terms of education, about one third (38.4%) of the respondents had a high school degree, and another third (35.8%) had less than a high school degree. Most respondents were not working at the time of their initial interview. Slightly less than half of the respondents (42.4%) identified the major source of income as legal work, and about one fifth (20.4%) stated that public assis- tance was their major source of income.

Table 2 describes participant substance use. The majority (51.1%) iden- tified crack or cocaine as their primary drug problem, followed by heroin and other opiates (20.6%), alcohol (12.2%), marijuana (3.0%), and amphet- amines (2.4%).

Theories of Motivation in Addiction Treatment 169

The average number of prior treatments for substance use disorders was 1.95 (SD = 4.4). Slightly less than half of the participants (44.7%) indi- cated that this was their first treatment experience for substance misuse. About one in five persons indicated that this was their second treatment experi- ence (20.5%) with the remaining participants reporting between three and six prior treatment episodes. About one third (34.7%) of the participants identified their primary referral source as self, followed by family or friends (30.9%), and the legal system (21.9%).

TABLE 1 Sample Characteristics

Variable N %

Gender Male 2,915 67.1 Female 1,432 32.9

Ethnicity African American 2,029 46.7 White 1,684 38.7 Hispanic 517 11.9 Other 117 2.7

Marital status at intake Never married 1,974 45.4 Married or living as married 1,374 31.6 Previously married 992 22.8 Missing 7 0.2

Age at admission 18–20 124 2.9 21–25 656 15.1 26–30 1,033 23.8 31–35 1,147 26.4 36–44 1,117 25.7 44 + 270 6.2

Educational level No high school degree 1,555 35.8 High school degree 1,670 38.4 Some college 741 17.0 College degree 377 8.7 Missing 4 0.1

Major source of income Legal work 1,843 42.4 Public assistance 887 20.4 Illegal sources 642 14.8 Family or friends 291 6.7 Social Security 198 4.6 No income 198 4.6 Other 61 1.4 Missing 227 5.2

Criminal justice status No current legal status 2,389 55.0 Current legal status 1948 44.8 Missing 10 0.2

170 K. Kennedy and T. K. Gregoire

We assessed problem severity by considering frequency of substance use and reports of regular use of more than one substance. Almost half of the respondents (48.4%) indicated that they used their primary drug daily, almost every day, or on multiple occasions in a single day. An additional 35.5% reported using one to six times per week, 3.3% reported using less than once a week, and 4.6% reported no use at the time of their initial inter- view. The number of different drugs used weekly ranged from zero to eight.

Measures and Procedure

To control for the role of problem severity we created a severity measure and categorized respondents into one of three severity groups. Participants were assigned to a low-severity group when the frequency of recent sub- stance use was reported as none or less than one time per week and the number of lifetime substances used was zero or one. Moderate-severity

TABLE 2 Substance Use and Treatment History

Primary drug problem N %

Cocaine/crack 2,221 51.1 Opiates 894 20.6 Alcohol 530 12.2 Other/not specified 374 8.6 Marijuana 129 3.0 Amphetamines 105 2.4 Missing 94 2.2

Treatment modality Short-term inpatient 1,379 31.7 Residential 1,193 27.4 Outpatient drug free 1,098 25.3 Methadone maintenance 677 15.6

Previous treatment No prior treatment 1,943 44.7 One prior treatment 893 20.5 Two or more prior treatments 1,499 34.5 Missing 12 0.3

Referral source Self 1,509 34.7 Family or friends 1,344 30.9 Legal 950 21.9 Other professional 535 12.3 Missing 9 0.2

Frequency of primary drug use 1–6 times per week 1,543 35.5 2 + times per day 1,118 25.7 Daily or almost every day 987 22.7 Less than once per week 145 3.3 None 202 4.6 Missing 352 8.1

Theories of Motivation in Addiction Treatment 171

participants reported frequency of use of less than one time per week and the use of three or more substances, or for those who reported daily use, of one or more substances. Persons who reported daily or more than daily use of one or two substances were also assigned to moderate severity. The high-severity groups included those reporting daily or more than daily use who also reported using three or more substances. This group also included people who reported using one to five times per week who were also using four or more substances. Determining severity in this manner resulted in 13.5% in the low-severity group, 57.3% in the moderate group, and the remaining 29.2% in the high-severity group.

To measure the independent variable of type of motivation (internal or external), we recoded responses to the question “What is the most impor- tant reason you are in treatment?” into three categories: internal, external, or unidentified. There were 57 possible responses for each of the questions: Responses were recorded as open-ended questions and then coded into a fixed category. Examples of external motivation include drug availability, custody issues with children, and court. Examples of internal motivation include disgusted with lifestyle, fear, and wanting to get off drugs. Unidentified responses were those that could have fallen into either internal or external motivation, such as religious reasons. A total of 47 of the possible responses were coded as external, 7 of the possible responses were coded as internal, and 3 of the possible responses were coded as undecided.

The survey repeated the important reason question three times to identify the primary, secondary, and tertiary reason for the client’s referral. Based on the responses to the three questions, a respondent could have endorsed three external responses, three internal responses, or a combination of the two types. Responses that were considered external were recoded as a neg- ative number. Responses were summed from the recoded responses for the three questions to create a motivation continuum, a scale ranging from –3 (totally external) to +3 (totally internal).

Respondents were initially categorized into six categories based on the source of motivation, ranging from –3 (totally extrinsic) to +3 (totally intrinsic). Scores of –2 and +2 were associated with somewhat external and somewhat internal motivation, respectively. Because of the low number of respondents in the somewhat external (n = 2) and somewhat internal categories (n = 4), these categories were collapsed under the lower heading. Somewhat inter- nal respondents were included with the slightly internal respondents, and somewhat external respondents were included with the slightly external respondents. The final source of motivation variable contained four levels: totally external, slightly external, totally internal, and slightly internal.

Approximately one third of the respondents had a motivation score of +1, indicating slight internal motivation (34.9%) and a similar percentage, with a motivation score of –1, were coded as slightly externally motivated (37.3%). Only 5.8% of the respondents had a score of –3, indicating totally

172 K. Kennedy and T. K. Gregoire

external motivation, and 21.4% scored +3 on the measure, indicating a totally internal source of motivation.

The DATOS study did not assess stage of change until the 3-month interview. At this point, the sample size was dramatically reduced (n = 3,180) both due to sample selection and attrition. Having participated in treatment for 3 months, the vast majority of the respondents were observed to be in the action stage. Given our goal of exploring the relationship between moti- vation type and stage of change at treatment inception, it was not a viable option to employ a sample based on 3 months of treatment participation and 30% attrition. Further, in light of the strong literature support that this attrition might be biased against earlier stages of change, we saw the need to create a measure of change at the outset of treatment (Callaghan et al., 2005; Edens & Willoughby, 2000; Haller et al., 2004; Simpson & Joe, 1993).

We created a proxy for stage of change from data available in the two interviews conducted within 1 week of treatment admission. The second of the two initial interviews included six questions from the Circumstances, Motivation, Readiness, and Suitability Scale (CMRS). The CMRS has been shown to have high internal consistency reliability and, as a measure of validity, is effective in predicting treatment retention (DeLeon, Melnick, Kressel, & Jainchill, 1994).

Our proxy measures employed responses to the six CMRS questions to assign clients to one of three mutually exclusive stages: precontemplation, contemplation, or action stage. The basis for assigning a stage was determined by whether a participant recognized a substance use problem, and whether they had taken any action in the past month to address that problem.

If a participant responded not at all to the statement “My drug use is a very serious problem in my life,” and responded very much agree to the statement “I don’t really need treatment, I’m here because of pressure on me,” we categorized them into the precontemplation stage. If a participant responded agree somewhat or very much agree to the four questions, “I feel that my drug use and the way I’ve been living have hurt a lot of people,” “I am really tired of using drugs and want to change,” “I really do need to be completely drug free in order to live the way I want to,” and “My drug use is a very serious problem in my life,” they were categorized into the con- templation stage.

We classified persons into the action stage based on their problem rec- ognition and whether they had engaged in change-oriented behavior in the 30 days prior to the interview. We employed each participant’s response to an inquiry about their attendance at self-help group meetings in the past 30 days or whether they had participated in any other treatment in the past 30 days. Affirmative responses to this question resulted in the participant being categorized into the action stage.

Slightly less than two thirds of the respondents (65.6%) were in the contemplation stage, one third were in the action stage (32.9%) and the

Theories of Motivation in Addiction Treatment 173

remaining 1.5% were in the precontemplation stage. Our approach to classi- fication represents what Migneault, Adams, and Read (2005) referred to as an algorithmic approach, in which individuals are classified into mutually exclusive stages of change based on specific responses to their current sub- stance use history. This approach has been used successfully in prior studies to classify individuals with substance use problems (Belding, Iguchi, Lamb, & Lakin, 1995; Migneault, Pallonen, & Velicer, 1997).

RESULTS

We began the process of model building by analyzing a series of bivariate relationships between the dependent variable stage of change and a num- ber of demographic and substance use variables that might be related to it. The purpose of the procedure was to develop a model that controlled for alternative explanations of variability in the dependent variable.

Table 3 describes the bivariate relationship of motivation source and the stage of change. Persons with an external source of motivation were more likely to be in the precontemplation stage. High internal source of motivation was associated with membership in the action stage. This rela- tionship was significant, c2(6, N = 4,347) = 54.385, p < 001, v = .079.

Our next analysis consisted of a series of cross-tabulations between stage of change and demographic variables described in Table 4. Because the chi-square statistic is dependent on sample size, our large sample size was likely to result in significant chi-square statistics despite trivial differ- ences between the stage of change groups. To address this concern we added a measure of effect size as an additional criterion for inclusion into the model. For these cross-tab analyses, we employed the Cramer’s v statistic using Cohen’s (1988) criteria of .10 for a small effect size.

Table 4 describes these bivariate relationships. Using the criteria of sta- tistical significance and a small effect size, only legal involvement (p < .001, v = .102) was significantly related to stage of change. No other demographic variable met the criteria for significance.

Our model-building procedure also considered a number of treatment and substance use variables that might be expected to impact stage of change.

TABLE 3 Percent Distribution of Motivation Source by Stage of Change

Precontemplationa Contemplationb Actionc

Total externald 15.9 5.4 6.1 Slight externale 68.1 36.0 36.8 Slight internalf 13.0 37.0 35.3 Total internalg 2.9 21.6 21.8

an = 69; bn = 2,852; cn = 1,426; dn = 252; en = 1,599; fn = 1,566; gn = 930.

174 K. Kennedy and T. K. Gregoire

There were five drug and alcohol use indicators that met inclusion criteria: referral source (p < .001, v = .137), drug of choice (p < .001, v = .222), and the number of prior treatments (p < .001, v = .120). Although severity did not meet the inclusion criteria of v > .1, we included it in the model due to the conceptual importance of the indicator (p < .001, v = .073). Because of the extensive literature on gender differences in treatment, we initially included this variable despite its lack of significance in the bivariate model (p = .071, v = .035). Gender remained nonsignificant in the multivariable model and was excluded from further analysis.

We then conducted a multinomial logistic regression to assess the influ- ence of source of motivation on stage of change. Stage of change was regressed on the following variables: source of motivation, drug of choice, treatment modality, legal status, educational level, referral source, number of prior drug treatments, and severity.

Our initial analyses were somewhat stymied by problems of quasi- separation in the data. This problem is caused by zero cell counts that occur when the dependent variable does not vary for some values of the categorical independent variables (Menard, 2002). This was not an unexpected problem given both the number of categorical independent variables in the initial analysis and the small number of people in the precontemplation group relative to the other two stage of change groups. Menard (2002) notes that although this problem does not impact the overall fit of a model, it does result in higher standard errors that will influence the individual regression coefficients. We took a twofold approach to addressing this problem: drop- ping the variables of educational status, referral source, and modality when preliminary results suggested their bivariate relationships to the dependent

TABLE 4 Bivariate Relationship with Stage of Change

Variable c2 df V

Demographic information Gender 5.29 2 .035 Age group 27.24* 10 .056 Ethnicity 28.45* 6 .057 Marital status 24.77* 4 .053 Educational level 8.77 6 .032 Major source of income 8.41 12 .025

Criminal justice status 45.10* 2 .102

Alcohol and drug indicators Primary drug 210.30* 10 .222 Treatment modality 24.77* 4 .053 Referral source 162.19* 8 .137 Previous treatment 124.21* 4 .120

Severity indicators Severity 42.70* 4 .073

*p < . 05.

Theories of Motivation in Addiction Treatment 175

variable were eliminated in this multivariable analysis; and collapsing the two categorical variables of prior treatment and drug of choice. The number of persons with three or more prior treatments was merged with those with two prior treatments, and persons reporting sedative (0.8% of sample), hallucinogenic (1.5%), or inhalant (0.1%) drugs of choice were merged into the other drug use category.

With the exclusion of cases with missing values, a total of 4,347 cases were included in this analysis. Multinomial logistic regression uses the chi- square as a test of the model’s significance. The full model –2 log likelihood test for source of motivation was significant, c2(26, N = 4,357) = 320.85, p < .001, indicating that the specified model was a significant improvement over a constant only, or null model. Taken as a measure of effect size, the Nagelkerke pseudo-R2 value of .104 represented a relatively small effect for the full model. The –2 log likelihood ratio test for source of motivation was signifi- cant, c2(6, N = 4,357) = 20.57, p = .002, indicating that a client’s source of motivation was significantly related to his or her stage of change.

Parameter estimates are created in this approach by making pairwise comparisons of each categorical outcome to a reference group. We employed the action stage of change as the reference category and compared those in the action stage with persons in the precontemplation or contemplation stages. By repeating the analysis with the precontemplation as the reference group we were also able to conduct a comparison of precontemplation to contemplation. Table 5 provides the results of these two analyses.

In addition to designating a reference group for the dependent variable, it is also necessary to designate reference groups for categorical independent variables. The totally internal source of motivation is the reference group in this analysis. The analysis makes comparisons of totally internally motivated persons to somewhat internally motivated, and to both the externally moti- vated categories. Reference groups for other variables are identifiable in Table 5 by their lack of parameter estimates.

Precontemplation Versus Action

External source of motivation was associated with an increased likelihood of being in the precontemplation rather than the action group. Persons who scored in the totally external source of motivation group were 15.7 times more likely than totally internal persons to be in the precontemplation rather than the action stage. Those scoring in the slightly external range were 10.1 times more likely to be in the precontemplation group than the action group. Being slightly internally motivated did not predict precontemplation versus action group membership.

The probability of being in the precontemplation rather than action group appeared to be greater with a decline in substance use severity, as persons in the lower severity category were four times more likely to be

176 K. Kennedy and T. K. Gregoire

TABLE 5 Parameter Estimates of Multinomial Regression Predicting Stage of Change from Motivation Source

B SE B df Odds ratio

Precontemplation vs. action

Motivation source Total external 2.754 1.106 1 15.698* Slight external 2.311 1.029 1 10.081* Slight internal 0.957 1.103 1 2.603 Total internal

Substance use severity Low severity 1.390 0.591 1 4.017* Moderate severity 0.460 0.572 1 1.584 High severity

Legal status No legal status –0.511 0.365 1 0.600 Current legal status

Number of prior treatments No prior treatment 2.968 1.035 1 19.448* One prior treatment 2.974 1.055 1 19.562* Two or more treatments

Drug of choice Alcohol 1.326 1.087 1 3.766 Marijuana 3.143 1.104 1 23.177* Cocaine 0.356 1.076 1 1.428 Opiate 0.698 1.267 1 2.011 Amphetamine 1.508 1.200 1 4.517 Other

Contemplation vs. action Motivation source

Total external 0.017 0.168 1 1.017 Slight external 0.037 0.094 1 1.038 Slight internal 0.085 0.094 1 1.089 Total internal

Substance use severity Low severity –0.272 0.116 1 0.762* Moderate severity –0.035 0.082 1 0.966 High severity

Legal status No legal status 0.290 0.071 1 1.336* Current legal status

Number of prior treatments No prior treatment 0.890 0.083 1 2.436 One prior treatment 0.482 0.096 1 1.619 Two or more treatments

Drug of choice Alcohol 0.004 0.218 1 1.004 Marijuana 0.700 0.291 1 2.014 Cocaine 0.346 0.201 1 1.414 Opiate 1.060 0.214 1 2.886* Amphetamine 0.145 0.288 1 1.155 Other

(Continued )

Theories of Motivation in Addiction Treatment 177

found in the precontemplation group as compared to those in the high- severity group. Persons in the no prior and one prior treatment group were more likely than those in the multiple prior treatment group to be identified as in precontemplation. Marijuana users were also more likely to be in the precontemplation stage of change.

Contemplation Versus Action

The source of motivation did not distinguish persons in the contemplation stage from those in the action stage of change. Low problem severity was associated with a reduced likelihood of being in the contemplation group (OR = .762). Persons with no current legal involvement were more likely to be in the contemplation group than the action group. As with the prior comparison, the odds of being in the contemplation group declined as the number of treatments increased. Those with no treatment history were 2.4 times more likely to be in the contemplation group and those with a single treatment episode were 1.6 times more likely to be in this group than in the action group. Opiate and marijuana users were also more likely to be in the contemplation rather than the action group.

TABLE 5 (Continued)

B SE B df Odds ratio

Precontemplation vs.contemplation Motivation source

Total external 2.737 1.102 1 15.439* Slight external 2.274 1.027 1 9.714* Slight internal 0.872 1.101 1 2.319 Total internal

Substance use severity Low severity 1.662 0.587 1 5.271* Moderate severity 0.495 0.570 1 1.640 High severity

Legal status No legal status –0.801 0.362 1 .449* Current legal status

Number of prior treatments No prior treatment 2.077 1.034 1 7.983* One prior treatment 2.492 1.054 12.079* Two or more treatments

Drug of choice Alcohol 1.322 1.085 1 3.751 Marijuana 2.443 1.092 11.507* Cocaine 0.010 1.073 1 1.010 Opiate –0.362 1.263 1 .697 Amphetamine 1.363 1.194 1 3.910 Other

Note. N = 4,347. *p < .05.

178 K. Kennedy and T. K. Gregoire

Precontemplation Versus Contemplation

External motivation was associated with an increased likelihood of being in the precontemplation rather than the contemplation group. Totally exter- nally motivated individuals were 15.4 times more likely to be in the precon- templation stage, and those with slightly external motivation were 9.7 times more likely to be in the precontemplation than the contemplation group. Internal motivation did not distinguish these two groups.

Persons with the lowest level of severity were 5.3 times more likely to be a member of the precontemplation group (OR = 9.8 for low severity and 2.6 for moderate severity). Having no current legal status was associated with a reduced likelihood of being in the precontemplation group. As with earlier comparisons, a lack of prior treatment was associated with an increased likelihood of being in the precontemplation rather than the contemplation group. Only marijuana use was significantly associated with stage of change when contrasting the contemplation and precontemplation stages. Persons who reported marijuana as their drug of choice were 11.5 times more likely to be in the precontemplation group.

DISCUSSION

Our primary purpose was to examine the relationship between SDT and the TTM. We used three mutually exclusive stages of change (precontemplation, contemplation, and action) to represent the TTM and the client’s source of motivation (internal and external) to represent SDT. Even when controlling for a number of other factors that might be expected to influence stage of change, the source of motivation, in some cases, did predict stage of change. Specifically, external motivation was associated with an increased likelihood of being in the precontemplation stage of change when comparing persons in that stage of change to either the contemplation or action stage of change. Motivation type did not distinguish persons who were contemplators from those in the action group.

Because people in the precontemplation stage do not view their sub- stance use as a problem (Prochaska et al., 1994), SDT would suggest that persons in this stage are likely to present for treatment more as a function of external pressures. Our findings were consistent with that hypothesis. Among persons in the precontemplation stage, only 13.2% presented to treatment with any internal source of motivation. Further, as the contemplation stage is characterized by having some recognition of a problem but experiencing ambivalence about change, SDT would suggest that persons in the contem- plation stage would have a higher degree of internal motivation than persons in the precontemplation stage, although still being somewhat influenced by external factors. The data appeared to support such a conclusion. Finally,

Theories of Motivation in Addiction Treatment 179

because people in the action stage are characterized by having taken some action, and hence having initiated a meaningful commitment to change, SDT would predict the highest level of internal motivation to occur in the action stage. Our findings were consistent with each of these suppositions. Even when controlling for other potentially motivating variables, lower levels of external motivation were associated with a higher stage of change in pair- wise comparisons of persons in the precontemplation stage with those in both the contemplation and action stages. The findings here are also consistent with other literature that suggests that the source of motivation is a determinant of the stage of change (O’Hare, 1996).

The significance of the relationship of the source of motivation to the stage of change is reflected in the definitions of each stage of change. Persons in the precontemplation stage do not have awareness that there is a problem, so it stands to reason that people in this stage would have an outside influ- ence prompting them into treatment. People in the action stage have taken steps toward behavior change and, by definition, have engaged in an activity to change the behavior in the last 30 days. In our study this action stage appears to be concurrent with adopting a less external rationale for change.

The TTM model is a widely known approach to considering how per- sons with addiction disorders change. Practitioners are well acquainted with this approach to change, and there is an articulated framework for interven- tion based on a client’s stage of change (DiClemente, 2003; Prochaska et al., 1992). Employing an SDT approach to differentiating sources of motivation and recognizing the contribution of motivation source toward the stage of change would seem a useful practice principle. At least in this study, it seems likely that adopting an internally motivated approach to change might be a prerequisite for progressing toward taking action in the addictions. To that end, this attempt at integrating these two theories appears to have some utility for practice.

The relationship of motivation source to stage of change appeared to occur independently of legal pressure to change. Although many individuals experience some type of coercive pressure to enroll in treatment, a client’s source of motivation cannot be ascertained simply by noting the referral source or precipitating event. Certainly in our data, source of motivation operated independently of a client’s legal status. As Marlowe et al. (1996) found, the client’s definition of pressure to enter treatment was much different from the referral source.

The nature of the relationship of the control variables to stage of change tended to make intuitive sense and could be construed to support our conceptualization of internal and external motivation. Being in an earlier stage of change was associated with an increased likelihood of having a less severe problem and fewer prior treatment experiences. Use of marijuana as a primary drug of choice was associated with an earlier stage of change. This substance is typically associated with less overt consequences for

180 K. Kennedy and T. K. Gregoire

chronic users than the other drugs of choice. As a result, and regardless of objective measures of dysfunction, these users might be less likely to acknowl- edge the need for behavior change. Current legal involvement was associ- ated with an increased probability of being in the action rather than the contemplation group. This is consistent with literature that found that legal coercion was associated with an increased likelihood of being in the action stage of change (Gregoire & Burke, 2004). Yet it is noteworthy that persons in the action stage, despite being more likely to have legal involvement, were also more likely to articulate an internal rationale for change.

Limitations of This Study

Among the limitations of this study were the variables created to reflect stage of change. Because of the timing of the creation of the stage of change variable in the original data, we were compelled to create proxy measures. Although the variables we created appear to be conceptually related to the constructs, they are unique to our work and lack the benefit of having been validated in prior research. Further, our approach to creating the action stage did not allow us to distinguish between persons who were participating of their own volition versus those being compelled to treat- ment. However, particularly in the addictions, few persons initiate a treat- ment activity without some type of encouragement from others and we do not find in the TTM literature a requirement that action stage behavior be self-initiated. The TTM acknowledges as much in describing the context of change and noting the influence of social systems, interpersonal relation- ships, and current life situations on stage of change (DiClemente, 2003).

As we have noted, the large sample size increased the likelihood of making significant findings in most of the analyses. We attempted to account for this by using a measure of effect size, but it is important to rec- ognize that all of the effects we observed here, although significant, were rather small. Finally, this study was conducted with secondary data, and as a correlational design our findings cannot be construed to imply causality. Further, Sutton (2001) cautioned against drawing causal inferences with TTM models created with cross-sectional data such as ours, noting that such models make assumptions about the differential influence of variables at each stage.

CONCLUSION

The purpose of this study was to determine the nature of the relationship between SDT and the TTM. Clinically, the findings have implications for treatment providers. Integrating components of both theories to better understand the client’s motivation for treatment could allow for a more

Theories of Motivation in Addiction Treatment 181

comprehensive understanding of the client at assessment. Treatment can also be tailored based on the source of motivation and the stage of change to help a client progress successfully through treatment.

Our preliminary effort to explore the relationship between these two theories has led us to tentatively conclude that they are complementary, and can be incorporated into a greater understanding of motivation in substance abuse. A logical next step is to assess the impact of source of motivation and stage of change on treatment outcomes. The impact of external motiva- tors on treatment outcomes tends to produced mixed results (Farabee, Prendergast, & Anglin, 1999). Using SDT to articulate source of motivation, while considering the role of external coercion, could make a contribution to understanding the role of coercion, motivation, and subsequent treatment outcomes.

REFERENCES

Bandura, A. (1997). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191–215.

Belding, M., Iguchi, M., & Lamb, R. (1997). Stages and processes of change as pre- dictors of drug use among methadone maintenance patients. Experimental and Clinical Psychopharmacology, 5, 65–73.

Belding, M., Iguchi, M., Lamb, R., & Lakin M. (1995). Stages and process of change among poly-drug users in methadone maintenance treatment. Drug and Alcohol Dependence, 39, 45–53.

Blanchard, K., Morgenstern, J., Morgan, T., Labouvie, E., & Bux, D. (2003). Motiva- tional subtypes and continuous measures of readiness for change: Concurrent and predictive validity. Psychology of Addictive Behaviors, 17, 56–65.

Burke, A., & Gregoire, T. (2007). Substance abuse treatment outcomes for coerced and noncoerced clients. Health and Social Work, 31, 7–15.

Callaghan, R., Hathaway, A., Cunningham, J., Vettese, L., Wyatt, S., & Taylor, L. (2005). Does stage-of-change predict dropout in a culturally diverse sample of adolescents admitted to inpatient substance-abuse treatment? A test of the tran- stheoretical model. Addictive Behaviors, 30, 1834–1847.

Center for Substance Abuse Treatment. (1999). Enhancing motivation to change in sub- stance abuse treatment (Treatment Improvement Protocol Series, No. 35, DHHS Pub No. [SMA] 00-3460). Washington, DC: U.S. Government Printing Office.

Clancy, J. (1961). Procrastination: A defense against sobriety. Quarterly Journal of Studies on Alcohol, 22, 269–276.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Davidson, R. (1998). The transtheoretical model: A critical overview. In W. R. Miller & N. Heather (Eds.), Treating addictive behaviors (2nd ed., pp. 25–38). New York: Plenum.

Deci, E., & Ryan, R. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum.

182 K. Kennedy and T. K. Gregoire

DeLeon, G., Melnick, G., Kressel, D., & Jainchill, N. (1994). Circumstances, motivation, readiness and suitability (The CMRS Scales): Predicting retention in therapeutic community treatment. American Journal of Drug and Alcohol Abuse, 20, 495–515.

DiClemente, C. (2003). Addiction and change: How addictions develop and addicted people recover. New York: Guilford.

DiClemente, C., & Hughes, S. (1990). Stages of change profiles in outpatient alco- holism treatment. Journal of Substance Abuse, 2, 217–235.

Edens, J., & Willoughby, F. (2000). Motivational patterns of alcohol dependent patients: A replication. Psychology of Addictive Behaviors, 14, 397–400.

Farabee, D., Prendergast, M., & Anglin, M. (1999). The effectiveness of coerced treatment for drug abusing offenders. Federal Probation, 62, 3–11.

Gregoire, T., & Burke, A. C. (2004). The relationship of legal coercion to readiness to change among adults with alcohol and other drug problems. Journal of Substance Abuse Treatment, 26, 337–343.

Haller, D., Miles, D., & Cropsey, K. (2004). Smoking stage of change is associated with retention in a smoke-free residential drug treatment program for women. Addictive Behaviors, 29, 1265–1270.

Heather, N., Rollnick, S., & Bell, A. (1993). Predictive validity of the Readiness to Change questionnaire. Addiction, 88, 1667–1677.

Henderson, M., Galen, L., & Saules, K. (2004). The predictive validity of the University of Rhode Island Assessment Questionnaire in a heroin-addicted polysubstance abuse sample. Psychology of Addictive Behaviors, 18, 106–112.

Marlowe, D., Kirby, K., Bonieskie, L., Glass, D., Dodds, L., Husband, S., et al. (1996). Assessment of coercive and non-coercive pressure to enter drug abuse treatment. Drug and Alcohol Dependence, 42, 77–84.

Menard, S. (2002). Applied logistic regression analysis (2nd ed.). Thousand Oaks, CA: Sage.

Migneault, J., Adams, T., & Read, J. (2005). Application of the transtheoretical model to substance abuse: Historical development and future directions. Drug and Alcohol Review, 24, 437–448.

Migneault, J., Pallonen, U., & Velicer, W. (1997). Decisional balance and stage of change for adolescent drinking. Addictive Behavior, 22, 339–351.

Miller, W., & Rollnick, S. (2002). Motivational interviewing: Preparing people for change (2nd ed.). New York: Guilford.

O’Hare, T. (1996). Court-ordered versus voluntary clients: Problem differences and readiness for change. Social Work, 41, 417–422.

Prochaska, J. (1979). Systems of psychotherapy: A transtheoretical analysis. Homewood, IL: Dorsey.

Prochaska, J., & DiClemente, C. (1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51, 390–395.

Prochaska, J., DiClemente, C., & Norcross, J. (1992). In search of how people change. American Psychologist, 47, 1102–1114.

Prochaska, J., Velicer, W., Rossi, J., Goldstein, M., Marcus, B., Rakowski, W., et al. (1994). Stages of change and decision balance for 12 problem behaviors. Health Psychology, 13, 39–46.

Theories of Motivation in Addiction Treatment 183

Ryan, R., Plant, R., & O’Malley, S. (1995). Initial motivations for alcohol treatment: Relations with patient characteristics, treatment involvement, and dropout. Addictive Behaviors, 20, 279–297.

Simpson, D., & Joe, G. (1993). Motivation as a predictor of early dropout from drug abuse treatment. Psychotherapy, 30, 357–368.

Sutton, S. (2001). Back to the drawing board? A review of applications of the trans- theoretical model to substance use. Addiction, 96, 175–187.

U.S. Department of Health and Human Services, National Institute on Drug Abuse. (2004). Drug Abuse Treatment Outcome Study (DATOS), 1991–1994: United States [Computer file] (2nd ICPSR ed.). Ann Arbor, MI: Inter-university Consor- tium for Political and Social Research.

Weinstein, N., Rothman, A., & Sutton, S. (1998). Stage theories of health behavior: Conceptual and methodological issues. Health Psychology, 17, 290–299.

Zeldman, A., Ryan, R., & Fiscella, K. (2004). Motivation, autonomy support, and entity beliefs: Their role in methadone maintenance treatment. Journal of Social and Clinical Psychology, 23, 675–696.