Ersche, Karen D. PhD*,†; Turton, Abigail J. BSc*,†; Croudace, Tim PhD*; Štochl, Jan PhD*
Drug addiction is a chronically relapsing disorder, characterized by a compulsive drive to seek drugs and a loss of control over drug intake.1 Multiple lines of evidence indicate that drug addiction is associated with significant disruptions in brain systems underlying self-control.2,3 For example, substance-dependent individuals demonstrate significant impairment in the control of behavior,4–6 attention,7–9 and other cognitive functions.10,11 These neurocognitive impairments have a considerable impact on the effectiveness of treatment,12–14 but drug users often lack insight into the extent of their cognitive deficits.15,16 Their beliefs of being in control, rather than their actual abilities, seem to play an important role in their engagement in treatment and determine their efforts in maintaining drug abstinence.17 The assessment of control beliefs of drug-dependent individuals may provide useful information for clinicians concerned with the treatment of drug users and also for researchers investigating drug users’ impairments in executive function.
The psychological concept of locus of control (LOC) reflects the belief of being in control, that is, the degree to which a person feels that rewards in life are contingent on their own behavior, or by contrast, are controlled by other people or external forces.18 LOC was initially formulated as an unidimensional construct representing a single continuum from internal to external sources of control and is measured by Rotter’s internal-external (I-E) scale. On the I-E scale, a person’s generalized control beliefs are reflected by a single score ranging from highly internal to highly external.18 Internally controlled individuals believe that successes or failures in life are due to their own efforts and abilities, whereas those with an external sense of control believe that control is out of their hands and that outcomes in life are determined by forces such as other people, luck or fate.18
To account for the variability of control experiences in different contexts, the LOC concept has been adapted to specific domains such as health,19 work,20 and drinking behavior,21 to mention a few. The drink-related LOC scale (DRIE),21 for example, takes into consideration the distinctive situations that are associated with the LOC in alcohol dependence. As one may expect, alcohol-dependent individuals report having more external control beliefs in alcohol-related situations compared with recovering alcoholics, who score more “externally” on the DRIE than social drinkers.22 The DRIE, however, is not without critics, in particular because of its mixture of both personalized (eg, “Once I start drinking I cannot stop”) and general statements (eg, “If people want something badly enough, they can change their drinking behaviour”). Alcohol-dependent patients have shown to make an external-to-internal shift in scoring on the DRIE during treatment,23 but this change may be confounded by the fact that personalized items may not be sensitive when alcohol is no longer consumed. General statements about drinking might be more appropriate for measuring changes in control beliefs. The DRIE scale has been transferred to drug-related situations by exchanging the world “alcohol” with the word “drug,”24–26 producing similar results to prior studies on the DRIE in alcohol-dependent populations. However, the problem with personalized items has not been addressed by any of the drug-related scales.
The aim of the present study was to develop a brief instrument to assess control beliefs with regard to drug-taking behaviors [drug-related locus of control (DR-LOC)] using only general statements, which has 2 advantages: (i) control beliefs can be assessed in drug users independently from their current drug-taking habits; and (ii) control beliefs can be measured in both drug using and non–drug-using individuals. The latter was a shortcoming for previous DR-LOC measures that were not suitable for non-drug users. We hypothesized that drug-dependent individuals would ascribe more control over their drug use to external forces (ie, deferring responsibility to others or external events rather than to themselves) compared with people who have never excessively used drugs. We will further explore whether a professional contact with drug users influences peoples’ persceptions of LOC in a drug-related context.
MATERIALS AND METHODS
Sample, Instruments, and Procedures
A total of 592 volunteers living in the East Anglia region of the UK completed the DR-LOC questionnaire and the Rotter’s I-E scale.18 The overwhelming majority of volunteers were of white race (89%), most of them were male (61% men, 39% women), and the mean age was 35.1 years (SD±11.7).
Approximately half of the volunteers were currently enrolled in an outpatient treatment program for opiates, stimulants, and alcohol dependence (n=310). These individuals were recruited because of their problematic drug-taking history. Drug-dependent volunteers were approached at treatment services in the East Anglia region, including London (UK) by clinically trained staff from the Mental Health Research Network (http://www.mhrn.info/). The other half of the sample (n=282) were recruited through the healthy volunteer panel at the Behavioural and Clinical Neuroscience Institute and by word of mouth within the local community. The control volunteers reported no personal history of substance dependence. Informed consent was obtained in writing from all participants. The study was approved by the NHS Cambridgeshire 2 Research Ethics Committee and the University of Cambridge Psychology Research Ethics Committee and has been performed in accordance with the Ethical Standards laid down in the Declaration of Helsinki.
All volunteers completed the 2 aforementioned questionnaires and provided further information, including their age, gender, years of education, and ethnic group by self-report. The initial DR-LOC contained 25-item pairs presented in a forced-choice format. Each pairing contained 1 statement indicating an internal control belief (eg, “Those who are successful in getting off drugs are often the lucky ones”) and 1 statement indicative of an external control belief (eg, “Getting off drugs depends upon lots of effort and hard work; luck has nothing to do with it”). Volunteers were instructed to choose the statement in each pair that most accurately described their current beliefs. The questionnaire items were presented in a balanced manner. Individual items were binary scored as either 0 (internal) or 1 (external). Rotter’s I-E scale,18 which measures generalized control beliefs in a neutral context,18,27 was included for comparison with responses in a drug-related context. We decided a priori to remove the 3 items of the I-E scale referring to situations at school, as these 3 items have been found inappropriate for adult samples.28
Group comparisons on demographic data were performed using t tests and χ2 methods, as appropriate, using the Statistical Package for the Social Sciences 13 (SPSS Inc.).
The response data from the initial 25-item DR-LOC questionnaire (shown in the Supplemental Digital Content 1, http://links.lww.com/ADTT/A1) were evaluated separately using exploratory factor analysis for binary data using maximum likelihood estimation in modern psychometric software. We used 12 integration points per dimension to increase the accuracy of the model. A χ2 (likelihood ratio test) determined the number of factors. Whist items with low factor loadings (below 0.3) were removed, meaningful factors were retained and interpreted based on their LOC concepts and theory. For completeness, factorial validity of each DR-LOC item was assessed simultaneously by both confirmatory factor analysis and a 2-parameter logistic item response model (both assuming a normal distribution for the latent factor continuum). We then used a more constrained (1-parameter) logistic model for the estimation of latent factor scores for all respondents. Internal consistency coefficients (lower bound estimates of reliability, under a restrictive test model) was estimated using Cronbach’s alpha for each factor as well as the composite reliability of the whole scale (within generalizability theory framework). The latter provides a more valid reliability estimate for these data. Psychometric modeling was performed using Mplus,29 BILOG,30 and mGENOVA.31
The final version of the DR-LOC questionnaire, retained items that loaded on 1 of 2 factors; each of the factors were loaded by 8 items (Table 1). These factors and the Rotter’s I-E were subject to group comparisons. The total scores of the Rotter’s I-E and the DR-LOC were analyzed using analysis of covariance; gender was included as a covariate. For group comparisons on the 2 DR-LOC subscales, multivariate analysis of covariance was conducted, also adjusting for gender. The least significant differences test was used to examine post hoc differences amongst the groups (ie, drug-dependent volunteers, health professionals, non-dependent healthy volunteers). Pearson correlations were performed where appropriate. All tests were 2-tailed and a significance level of 0.05 was assumed.
Psychometric Properties and Factor Structure of the DR-LOC
The likelihood ratio test of the exploratory factor analysis results revealed that a 2-factor solution was superior to the single unidimensional solution (χ2=196.8, df=29, P<0.001). A 3-factor solution was not indicated over 2 factors (χ2=12.9, df=16, P=0.68). Interfactor correlation was moderate, at r=−0.39. The first factor (addiction recovery) contained 8 items, reflecting the extent to which an individual believes that recovery from addiction is determined by their personal efforts rather than by the support provided from other people. The second factor (drug-taking decisions) also contained 8 items and measured the extent to which a person believes that the decision to take drugs is under their own personal control rather than determined by peer pressure, or external needs or problems. Nine items did not load on any of the 2 factors (loadings <0.3) and therefore were excluded from the final scale.
After the exclusion of items with low factor loadings, we applied a more restrictive 1-parameter logistic model, to simplify the interpretation of raw scores and further justify the use of summary scores. The 1-parameter logistic model is a more restrictive model, but model comparisons showed a non-significant decrease of model fit for both DR-LOC factors [ie, the deviance test with 8 degrees of freedom equals to 8.7 (P=0.46) for addiction recovery, and to 10.7 (P=0.21) for drug-taking decisions]. Further, graphical inspection supported this simpler model. No gross indications of misfit between any of the individual items and either the empirical item characteristic curve or the modeled item characteristic curve (see also Figure S1) were identified. This provided further support for the use of summary scores within each factor.
Internal consistency of the DR-LOC, as estimated by Cronbach’s alpha, was relatively low for addiction recovery (α=0.48) and moderate for drug-taking decisions (α=0.60), reflecting the breadth of the construct measured by these items. Reliability of the whole scale, as estimated by composite Cronbach’s alpha (yield 0.62), which is quite low, but still comparable with LOC measures previously used in drug user samples.26 Although Cronbach’s alpha has been considered the gold standard of indexing reliability, it may be challenged when the number of items are low and when the sampling errors impact on the range of candidate ability. In the present study, we did not find evidence that the SEM was affected by these problems as the SEM yielded 1.06 for addiction recovery and 1.19 for drug-taking decisions. The DR-LOC may therefore be an appropriate measure and could be recommended for routine use, but only for comparing quite large groups (n>100). Further details about the psychometric properties can be obtained from the authors (K.D.E., J.S.).
Volunteers with and without a drug dependency problem did not significantly differ from each other in terms of age (t456.2=0.96, P=0.337) or ethnicity (χ2=2.67, P=0.102). There was, however, a significant difference in the gender ratio (χ2=49.22, P<0.001); while in the control group the ratio of men (46%) and women (54%) was almost equally balanced; the drug user group was predominantly male (75% men, 25% women). Consequently, all group comparisons are reported after adjustment for main effects of gender. Drug users reported significantly fewer years of education (mean: 10.4 y, SD±2.7) compared with control volunteers (mean: 12.7 y, SD±2.5) (t567=10.01, P<0.001); this is a common finding, which has also been reported previously.32 As years of education were not correlated with either the DR-LOC scores (r=0.003, P=0.946) or the I-E total scores (r=−0.01, P=0.789), education was not included as a covariate in further analyses.
Twenty percent of non-dependent volunteers (eg, nurses, key workers, general practitioners) reported having professional contact with chronic drug users. This subgroup of volunteers was slightly older (mean: 39.4 y, SD±12.1) than those without professional contact with drug users (mean: 34.6 y, SD±14.5) (t272=2.34, P=0.020). However, age was not correlated with DR-LOC (r=−0.0.03, P=0.676) or the I-E scale (r=0.0.07, P=0.226) within the group of non-dependent volunteers and therefore was not included as a covariate in further analyses. Neither years of education (t261=−4.14, P=0.680), nor the gender ratio (χ2=2.0, P=0.157) differed between these 2 subgroups.
Drug-related Control Orientation With Regard to Personal Experience of Addiction
The drug-dependent individuals reported significantly greater levels of internal control in a drug-related context (mean score: 8.8, SD±2.5) compared with non-dependent volunteers (mean: 9.3, SD±2.8; F1,577=5.09; P=0.024). This significant difference was due to drug users having an internal sense of control in terms of addiction recovery compared with the non-dependent volunteers (drug users: 6.1, SD±1.5; volunteers: 6.5, SD±1.4; F1,577=12.80; P<0.001). No group difference was observed on the DR-LOC decision-making subscale (drug users: 2.7, SD±1.8; volunteers: 2.7, SD±2.1; F1,577=0.13; P=0.719). Generalized control beliefs, as measured by Rotter’s I-E scale did not significantly differ between the groups (drug users: 9.2, SD±3.2; volunteers: 9.5, SD±3.6; F1,577=3.43; P=0.065). The inclusion of the Rotter’s I-E score as a covariate in the analysis of the DR-LOC subscales, to control for variances in generalized control beliefs, did not change the results on the DR-LOC. This is consistent with the notion that drug-related control beliefs are domain specific and not explained by general control beliefs. Yet, Rotter’s I-E scale and the DR-LOC were not independent, rather they weakly correlated with each other (r=0.27, P<0.001).
Experience With Drug-addicted Individuals and Drug-related Control Orientation
Comparisons between drug users and the 2 subgroups of non-dependent volunteers revealed a significant difference in drug-related control beliefs (F2,576=3.70, P=0.025), but not in terms of generalized control beliefs (F2,576=2.22, P=0.109). Group differences on the DR-LOC were reflected on both subscales. Thus beliefs of “being in control” when deciding to take drugs differed significantly between the groups (F2,576=5.19, P=0.006). As shown in Figure 1, health professionals were significantly more internally orientated when compared with their non-professional counterparts (P=0.001), and also when compared with the drug users (P=0.018). However, control beliefs between non-professional volunteers and the drug users did not differ (P=0.167). There was also a significant group difference in terms of control beliefs regarding addiction recovery (F2,576=7.45, P=0.001). Post hoc analysis revealed that drug users were significantly more internally controlled when separately compared with either health professionals (P=0.001) or nonprofessional volunteers (P=0.004), while the beliefs in the 2 non-dependent groups were similar (P=0.150). The significant group differences on both subscales and post hoc tests survived when the analysis was corrected for generalized LOC beliefs using the Rotter’s I-E score as a covariate.
The DR-LOC is a new instrument for the assessment of subjective control beliefs in a drug-related context. The questionnaire is short, consisting of 2 scales measuring individuals’ perceptions of control regarding (i) the decision to take drugs; and (ii) the attainment of addiction recovery. In this initial study, we have shown that self-reported drug-related control beliefs differ significantly depending on prior experience with drug addiction, and this variation in control beliefs was not seen on the generalized LOC measure. This is an important first step in the validation of this new instrument; however, further clinical validation is needed to demonstrate its ability to accurately score individuals and to distinguish among clinical groups of patients (eg, in treatment versus non-treatment seeking). These initial validation results suggest that the DR-LOC has greater predictive validity for drug-taking behaviors than the Rotter’s I-E scale, justifying the development of a drug-taking–specific adaptation of the Rotter’s I-E scale.
Personal or Professional Experiences With Addiction Modulate Drug-related Control Beliefs
Drug addiction is associated with significant experiences in losing control, but these experiences are not captured by generalized control beliefs, which may be the reason why the Rotter’s I-E scale does not show discriminative validity between drug-dependent and non-dependent individuals.33–35 The present study provides preliminary evidence that both professional and personal experiences with drug addiction affect how a person perceives control in a drug-related context. Our data show that drug-dependent individuals have a much more internal sense of control in terms of addiction recovery than nondependent individuals or people with professional experiences of addiction. It is, however, of note that all our drug-dependent volunteers were currently enrolled in a drug treatment program and in previous studies generalized control beliefs have been shown to change during treatment, from a more external toward a more internal orientation.36–38 It is conceivable that their receipt of treatment affected drug users’ judgments about addiction recovery.
Our data also showed that professional experiences with drug users had an impact on their drug-related control beliefs. This observation concurs with findings from previous studies showing that health professionals’ attitudes toward patients with and without drug-taking histories differ notably.39–42 As shown in Figure 1, the health professionals in our study considered external factors as critical for treatment success, whereas the drug-dependent individuals attributed to themselves the control over their treatment progress. A significant discrepancy also emerged with regard to decisions to take drugs: health professionals attributed control over drug-taking to internal sources to a significantly greater degree than the drug users. These discrepancies between drug users and health professionals are striking, but may not be unusual. Previous studies comparing the views of drug users and their therapists on treatment progress and drug-taking habits also found significant discrepancies.43 Furthermore, these studies revealed that the drug users’ perceptions on these issues had greater predictive value for the therapeutic outcome than those provided by their therapists.43,44 The predictive value of the DR-LOC scores for treatment outcome has to be established by future studies, but it is conceivable that the DR-LOC could help improve the therapeutic alliance between drug users and their counselors by highlighting the discrepancies in their control beliefs.
The DR-LOC may be regarded as an equivalent to the DRIE scale for a drug-related context. From a conceptual point of view, this comparison holds, but the item structure in the DR-LOC is different. The DRIE scale, which assesses drink-related control beliefs, includes personalized statements (eg, “I feel completely helpless when it comes to drinking”; or “Once I start drinking I cannot stop”), which are not suitable for individuals who never drink alcohol or those who have given up drinking alcohol. For the DR-LOC, we deliberately avoided personalized statements but used general statements instead (eg, “There are people who feel completely helpless when it comes to resisting taking drugs”), allowing the assessment of control beliefs in individuals irrespective of personal drug-taking experiences. This item structure is a key strength of the DR-LOC and distinguishes it from other LOC measures that were previously used in drug users.24,26 Furthermore, the DR-LOC allows one to measure all types of drug-related control beliefs as opposed to beliefs related to just 1 particular drug (ie, alcohol) because loss of control is a hallmark of addictive behavior in general and not specific to the drug of dependence. Moreover, the DR-LOC scale published here is a questionnaire measure with the potential to investigate environmental factors on control beliefs in individuals from very different backgrounds including, non-dependent family members, prison staff, police officers, or school teachers.
Psychometric Properties of the DR-LOC Compared With Other Measures of LOC
Our data are in keeping with previous studies suggesting that control beliefs are a multidimensional rather than unidimensional construct.45–47 Internal reliability of Rotter’s I-E scale has been deemed as low, but adequate, ranging from 0.65 to 0.79,18,46,47 whereas Cronbach’s alphas of the subscales seem to vary between 0.58 and 0.70 in different populations.28 We acknowledge that the Cronbach’s alphas estimate of 0.62 for the DR-LOC scale was lower than for the average I-E scale, but the low internal consistent reliability estimate should not be considered as a rationale for discarding data.48 The standard error (SE) of measurement and the confidence intervals provide more information than reliability estimates alone,48 and both proved sufficiently low (accurate scores) in our sample to justify its use in research comparing groups (using means). It is also of note that the DR-LOC contains only 16 items compared with 23 items in the I-E scale; a fact that is important to consider when Cronbach’s alpha estimates are based on the total number of items. Future revisions of the DR-LOC scale may therefore involve increasing the total number of items on the scale.
Potential of the DR-LOC
Generalized LOC is thought to be one of the most widely studied psychological constructs.49 The extension of the concept to drug abuse may offer new opportunities for both research and clinical practice. Not only is the loss of control central to the clinical phenotype of addiction, there also is growing evidence that the context in which control is exerted is relevant to drug-dependent individuals. For example, recent research has shown that the brain systems underlying inhibitory cognitive control are not measurably impaired in drug users when challenged in a neutral context, however, these systems become compromised when the neutral stimuli are replaced with drug-related cues.7 Importantly, in alcohol dependence it has recently been shown that compromised cognitive control in an alcohol-related context can improve with cognitive training.50 These cognitive improvements were concomitant with changes in drink-related control beliefs (as assessed using the DRIE) in patients with alcohol dependence.50 Whether or not changes in DR-LOC beliefs are equally amenable to treatment is an important question for future clinical research to address. Preliminary evidence indicates that therapeutic interventions in psychiatric patients are more successful when generalized control beliefs are taken into account.51,52
For neuroscientific research investigating impairments in executive control, the DR-LOC could also prove useful as an adjunct measure to better understand the neural correlates of control in an addicted population. Preliminary evidence suggests that generalized control beliefs are related to cortisol regulation and hippocampal volume in healthy volunteers,53 systems that are known to be altered in addiction.54,55 It has further been hypothesized that individual differences in generalized control beliefs may be mediated by dopaminergic neurotransmission.56,57 In light of the pivotal role of dopamine in drug addiction,58 a link with control beliefs in a drug-related context may be likely.
Limitations and Conclusions
We have presented a brief novel scale with demonstrated validity in measuring drug-related control beliefs in both drug-dependent and non-dependent individuals. When developing the measure, the use of the first person in the wording of the questionnaire items was deliberately avoided in an effort to allow the instrument to assess control beliefs across a range of individuals. In relation to this, it is important to note that we demonstrate that the DR-LOC has good discriminative value in differentiating the control beliefs of individuals with different prior experiences of drug addiction. The weakness of the DR-LOC is the relatively low internal consistency of the addiction recovery scale, despite the good factor loadings of items. However, this issue could be addressed in future studies using a revised DR-LOC scale.
The authors thank all participants for their contributions to this study and the National Institute of Health Research (NIHR) Mental Health Research Network (MHRN) for their dedicated support with data collection. Special thanks therefore go to Rachel Everett, Harriet McGrath, Muireann McSwiney, Jenny Lacey, Lauren Wright, Lucy Wigg, Lorna Jacobs, Mariam Errington, Naomi Bateman, Gabriel Abotsie, Regina Barreto, Zoe Given-Wilson, Kathryn Betts, Adrian Jackson, and Angela Browne. The authors also thank Paula Cruise for advice on questionnaire design, Anna Brown for statistical advice, and James Wason for comments on the final draft.
1. Diagnostic and Statistical Manual of Mental Disorders. 20004th ed, Text Revision Washington, DC American Psychiatric Association
2. Goldstein RZ, Volkow ND. Drug addiction and its underlying neurobiological basis: neuroimaging evidence for the involvement of the frontal cortex. Am J Psychiatry. 2002;159:1642–1652
3. Jentsch JD, Taylor JR. Impulsivity resulting from frontostriatal dysfunction in drug abuse: implications for the control of behavior by reward- related stimuli. Psychopharmacology. 1999;146:373–390
4. Garavan H, Hester R. The role of cognitive control in cocaine dependence. Neuropsychol Rev. 2007;17:337–345
5. Fillmore MT, Rush CR. Impaired inhibitory control of behavior in chronic cocaine users. Drug Alcohol Depend. 2002;66:265–273
6. Monterosso JR, Aron AR, Cordova X, et al. Deficits in response inhibition associated with chronic methamphetamine abuse. Drug Alcohol Depend. 2005;79:273–277
7. Ersche KD, Bullmore ET, Craig KJ, et al. Influence of compulsivity of drug abuse on dopaminergic modulation of attentional bias in stimulant dependence. Arch Gen Psychiatry. 2010;67:632–644
8. Marissen MAE, Franken IHA, Waters AJ, et al. Attentional bias predicts heroin relapse following treatment. Addiction. 2006;101:1306–1312
9. Carpenter KM, Schreiber E, Church S, et al. Drug stroop performance: relationships with primary substance of use and treatment outcome in a drug-dependent outpatient sample. Addic Behav. 2006;31:174–181
10. Ersche KD, Clark L, London M, et al. Profile of executive and memory function associated with amphetamine and opiate dependence. Neuropsychopharmacology. 2006;31:1036–1047
11. Verdejo-Garcia A, Bechara A, Recknor EC, et al. Executive dysfunction in substance dependent individuals during drug use and abstinence: an examination of the behavioral, cognitive and emotional correlates of addiction. J Int Neuropsychol Soc. 2006;12:405–415
12. Aharonovich E, Nunes E, Hasin D. Cognitive impairment, retention and abstinence among cocaine abusers in cognitive-behavioral treatment. Drug Alcohol Depend. 2003;71:207–211
13. Aharonovich E, Hasin DS, Brooks AC, et al. Cognitive deficits predict low treatment retention in cocaine dependent patients. Drug Alcohol Depend. 2006;81:313–322
14. Teichner G, Horner MD, Harvey RT. Neuropsychological predictors of the attainment of treatment objectives in substance abuse patients. Int J Neurosci. 2001;106:253–263
15. Goldstein RZ, Craig AD, Bechara A, et al. The neurocircuitry of impaired insight in drug addiction. Trends Cogn Sci. 2009;13:372–380
16. Verdejo-Garcia A, Perez-Garcia M. Substance abusers’ self-awareness of the neurobehavioral consequences of addiction. Psychiatry Res. 2008;158:172–180
17. Murphy PN, Bentall RP. Motivation to withdraw from Heroin—a factor-analytic study. Br J Addict. 1992;87:245–250
18. Rotter JB. Generalized expectancies for internal versus external control of reinforcement. Psychol Monogr. 1966;80:1–28
19. Wallston KA, Wallston BS, Devellis R. Development of multidimensional health locus of control (Mhlc) scales. Health Educ Monogr. 1978;6:160–170
20. Spector PE. Development of the work locus of control scale. J Occup Psychol. 1988;61:335–340
21. Donovan DM, Oleary MR. Drinking-related locus of control scale - reliability, factor structure and validity. J Stud Alcohol. 1978;39:759–784
22. Huckstadt A. Locus of control among alcoholics, recovering alcoholics, and non-alcoholics. Res Nurs Health. 1987;10:23–28
23. Abbott MW. Locus of control and treatment outcome in alcoholics. J Stud Alcohol. 1984;45:46–52
25. Hartmann DJ. Replication and extension analyzing the factor structure of locus of control scales for substance-abusing behaviors. Psychol Rep. 1999;84:277–287
26. Oswald LM, Walker GC, Krajewski KJ, et al. General and specific locus of control in cocaine abusers. J Subst Abuse. 1994;6:179–190
27. Coombs WN, Schroeder HE. Generalized locus of control: an analysis of factor analytic data. Personal Individ Diff. 1988;9:79–85
28. Cherlin A, Bourque LB. Dimensionality and reliability of Rotter I-E scale. Sociometry. 1974;37:565–582
29. Muthén LK, Muthén BO. Mplus: Statistical Analysis With Latent Variables (Version 6.1). Los Angeles; CA: (1998-2010)
30. Zimovski M, Muraki E, Mislevy R, et al. BILOG-MG (Version 3.0.2776.1). Chicago, IL: Scientific Software Inernational, 2007
31. Brennan RL. Manual for mGENOVA (Version 2.1). Iowa City, IA: University of Iowa, Iowa Testing Programs, 2001
32. Booth BM, Leukefeld C, Falck R, et al. Correlates of rural methamphetamine and cocaine users: results from a multistate community study. J Stud Alcohol. 2006;67:493–501
33. Doherty O, Matthews G. Personality characteristics of opiate addicts. Personal Individ Diff. 1988;9:171–172
34. Platt JJ. “Addiction proneness” and personality in heroin addicts. J Abnorm Psychol. 1975;84:303–306
35. Mitchell JM, Tavares VC, Fields HL, et al. Endogenous opioid blockade and impulsive responding in alcoholics and healthy controls. Neuropsychopharmacology. 2007;32:439–449
36. Lennings CJ. Changes in responses of heroin-addicts on the locus of control scale in a therapeutic-community—Odyssey House. Aust Psychol. 1980;15:359–367
37. Figurelli GA, Hartman BW, Kowalski FX. Assessment of change in scores on personal control orientation and use of drugs and alcohol of adolescents who participate in a cognitively oriented pretreatment intervention. Psychol Rep. 1994;75:939–944
38. De Leon G, Skodol A, Rosenthal MS. Phoenix house: changes in psychopathological signs of resident drug addicts. Arch Gen Psychiatry. 1973;28:131–135
39. Howard MO, Chung SS. Nurses’ attitudes toward substance misusers. II. Experiments and studies comparing nurses to other groups. Subst Use Misuse. 2000;35:503–532
40. Moodleykunnie T. Attitudes and perceptions of health-professionals toward substance use disorders and substance-dependent individuals. Int J Addict. 1988;23:469–475
41. Watson H, Maclaren W, Kerr S. Staff attitudes towards working with drug users: development of the Drug Problems Perceptions Questionnaire. Addiction. 2007;102:206–215
42. Knox WJ. Attitudes of psychologists toward drug-abusers. J Clin Psychol. 1976;32:179–188
43. Walton MA, Blow FC, Booth BM. A comparison of substance abuse patients’ and counselors’ perceptions of relapse risk: relationship to actual relapse. J Subst Abuse Treat. 2000;19:161–169
44. Barber JP, Luborsky L, Crits-Christoph P, et al. Therapeutic alliance as a predictor of outcome in treatment of cocaine dependence. Psychother Res. 1999;9:54–73
45. Parkes KR. Dimesionality of Rotter’s locus of control scale: an application of the ‘very simple structure’ technique. Personal Individ Diff. 1985;6:115–119
46. Harper H, Oei TPS, Mendalgio S, et al. Dimensionality, validity, and utility of the I-E scale with anxiety disorders. J Anxiety Disord. 1990;4:89–98
47. Lange RV, Tiggemann M. Dimensionality and reliability of the Rotter I-E locus of control scale. J Personal Assess. 1981;45:398
48. Onwuegbuzie AJ, Daniel LG. A framework for reporting and interpreting internal consistency reliability estimates. Meas Eval Counsel Dev. 2002;35:89–103
49. Judge TA, Erez A, Bono JE, et al. Are measures of self-esteem, neuroticism, locus of control, and generalized self-efficacy indicators of a common core construct? J Personal Soc Psychol. 2002;83:693–710
50. Fadardi JS, Cox WM. Reversing the sequence: reducing alcohol consumption by overcoming alcohol attentional bias. Drug Alcohol Depend. 2009;101:137–145
51. Abramowitz CV, Abramowitz SI, Roback HB, et al. Differential effectiveness of directive and nondirective group therapies as a function of client internal-external control. J Consult Clin Psychol. 1974;42:849–853
52. Snowden LR. Personality tailored covert sensitization of heroin abuse. Addict Behav. 1978;3:43–49
53. Pruessner JC, Baldwin MW, Dedovic K, et al. Self-esteem, locus of control, hippocampal volume, and cortisol regulation in young and old adulthood. NeuroImage. 2005;28:815–826
54. Koob GF, Le Moal M. Plasticity of reward neurocircuitry and the ‘dark side’ of drug addiction. Nat Neurosci. 2005;8:1442–1444
55. Nestler EJ. Common molecular and cellular substrates of addiction and memory. Neurobiol Learn Mem. 2002;78:637–647
56. De Brabander B, Declerck CH. A possible role of central dopamine metabolism associated with individual differences in locus of control. Personal Individ Diff. 2004;37:735–750
57. Declerck CH, Boone C, De Brabander B. On feeling in control: a biological theory for individual differences in control perception. Brain Cogn. 2006;62:143–176
58. Volkow ND, Fowler JS, Wang GJ, et al. Imaging dopamine’s role in drug abuse and addiction. Neuropharmacology. 2009;56(suppl 1):3–8
© 2012 Lippincott Williams & Wilkins, Inc.