K Austin The Efficacy Of Motivational Interviewing With Offenders


Published on

Published in: Health & Medicine, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Clients were often dismissed for not being motivated for programmes. Motivation was seen as the client’s problem not the therapist’s. Psychologists would attempt confrontational techniques to break through perceived “denial”, which perpetuating resistance. In the late 1970s this paradigm shifted following research that demonstrated clients’ intrapsychic characteristics played a minor role compared to therapist variables in predicting client motivation and treatment outcome. These studies redirected interest in the therapist’s ability, through the therapeutic relationship, to foster client motivation to change (Miller, 1985). Interest was stimulated in correctional rehabilitation where clients were viewed as lacking the requisite motivation to change their harmful behavior .
  • MI stemmed from experimental social psychology and other psychological theories, such as causal attributions (Weiner, 1986), cognitive dissonance (Festinger, 1957), and self-efficacy theory (Bandura, 1977). MI was explicated in the late 1970s and early 1980s as “a client-centered, directive method for enhancing intrinsic motivation to change by exploring and resolving ambivalence to change” (Miller & Rollnick, 2002, p. 25). MI has been applied to a range of problems, including decreasing risky sexual behaviour, alcohol use, drug use, diet and exercise, increasing adherence to medication and encouraging healthy lifestyle changes. The majority of literature on MI with offenders has involved recommendations rather than empirical studies. The popularity of MI has been, at least partially, driven by face validity (Czuchry, Sia, & Dansereau, 2006; Harper & Hardy, 2000). People often report that MI inherently makes sense and in turn presume its efficacy.
  • Risk posits that offenders who are more likely to re-offend benefit from more intensive interventions, while those who are less likely to re-offend benefit from less intensive interventions. Risk factors can be both static (unchangeable) and dynamic (changeable) (Andrews & Bonta, 2006; Cullen & Gendreau, 2000; McGuire, 2000). Needs are the malleable components of an offender’s risk of recidivism. The responsivity principle states that offenders benefit from a style and mode consistent with their ability and way of learning (Andrews & Bonta, 2006). An important component of responsivity is readiness to change, which includes motivation (Day & Howells, 2007; Day, Howells, Casey, Ward, & Birgden, 2007). An offender low in motivation to change is unlikely to benefit from action oriented interventions, because the individual is not at a stage of readiness to consider and act on new learning (Miller & Rollnick, 2002).
  • A RoC*RoI score of .7 indicates a 70% likelihood of re-offending within 5 years The principles provide the link between the gestalt spirit (collaboration, evocation and autonomy) of MI and in-session clinical skills (Miller & Rollnick, 1991). These principles include developing discrepancy, avoiding argumentation, rolling with resistance, expressing empathy and supporting self-efficacy. The skills of MI allow the clinician to translate the principles into clinical techniques. Each skill facilitates the client towards resolving ambivalence, building motivation and progressing towards behaviour change. This is primarily done by reinforcing change talk and building commitment to change.
  • A power analysis was used to calculate the required sample size to detect an effect. This was based on Anstiss’ (2005) comparable study with an effect size of eta squared = 0.27 ; a p value = 0.05; and a power level = 0.80 Index offences of burglary (75%), sex offences (9.4%), aggravated robbery (3.1%), drug offences (3.1%), driving offences (6.3%) and assault (3.1%) The sample included a disproportionate number of Māori, 76.3% compared to approximately 50% and younger offenders, M = 27.24 years of age compared to M = 31 years of age (Department of Corrections, 2003). However, research carried out by Wilson (2004) found that high risk offenders (defined as those who had a 70% chance of re-offending within five years) were disproportionately Māori, were on average younger and showed a pervasive pattern of criminal versatility. Independent samples t-tests were carried out to ascertain the comparability of the sub-group of 12 and the total sample. These showed that there were no significant differences in their motivation to change and risk scores, therefore the sub-group was considered representative.
  • The SMP was delivered individually within the prison over five weekly one-hour sessions by registered psychologists. The SMP manual (Steyn & Devereux, 2006) reflected the principles of effective correctional rehabilitation (Andrews & Bonta, 2006) and was delivered in concordance with the spirit, principles and skills of MI (Miller & Rollnick, 2002). SMP aimed to increase motivation to change so that upon release offenders would engage in social and community resources to address their rehabilitative needs.
  • 1. A statistically significant increase in offenders’ motivation to change scores from pre to post SMP. The maintenance of offenders’ post motivation to change scores at follow-up. This will in turn provide evidence for the efficacy of MI for high risk offenders (risk of recidivism scores of .7 and over). 2. Psychologists demonstrating MI skills which meet or exceed the MI Treatment Integrity Code 3.0 cut-off scores for ‘competency’.   3. Mean scores of four or greater for each principle on the A5PMI. This cut-off score represents a level at which offenders predominantly report strongly experiencing MI principles.
  • Quasi experimental – no control group. The URICA is a 32 item, structured self report questionnaire with four hypothesised factors (DiClemente & Hughes, 1990). Factor analysis has validated the four factors labelled pre-contemplation, contemplation, action and maintenance (McConnaughy, Prochaska, & Velicer, 1983). In the current study, Cronbach’s alpha coefficient was .80 for pre-contemplation, .70 for contemplation, .78 for action, .61 for maintenance and .65 for the SMP URICA total score. A measure of socially desirable responding was included, given the transparency of the SMP URICA and the Assessment of the Five Principles of Motivational Interviewing and feasible pressure on offenders to appear to have benefited from the SMP. Impression management captures the conscious use of inaccurate self-descriptions, such as malingering. Self-deceptive enhancement represents an unconscious process to deny psychologically threatening cognitions and affect. Cronbach’s alpha coefficients for this study were .47 for SDE, .85 for IM and .86 for the BIDR total score. The MITI has two components: global scores and behaviour counts (Moyers, Martin, Manuel, Miller, & Ernst, 2007). For every principle of MI two statements were developed. These were based on descriptions of the nature of the five principles in the literature (Miller & Rollnick, 2002). One statement reflected the individual’s response if the principle was adhered to and the other if the principle was contravened. The internal consistency of the A5PMI was found to be .71 using Cronbach’s alpha coefficient from the current study’s data.
  • Time 1 ( M = 129.61, SD = 13.23) to Time 2 ( M = 133.89, SD = 14.51), t (37) = 2.99, p < .05 (two tailed). For the sub-group, there was a significant effect for time with Wilks’ Lambda = .61, F (2, 10) = 3.15, p < .15. Multivariate partial eta squared of 0.39 indicated this to be a large effect size (Cohen, 1988). The statistically insignificant increase in motivation to change from post SMP to follow-up indicated that motivation was generally maintained at Post SMP levels
  • Psychologists only demonstrated full competency for direction, although global clinician ratings, percent open questions, evocation, collaboration, autonomy and empathy all closely approached competence. Psychologists received a rating of less than beginning proficiency for reflection to question ratio, complex reflections and MI adherent behaviours – these are skills specific to MI. The hypothesis that Psychologists would demonstrate competence in the use of MI skills was not fully supported.
  • An arbitrary cut-off score of four was used to indicate when a principle was strongly experienced. Other than “rolling with resistance”, offenders strongly experienced MI principles. Therefore, the hypothesis that offenders will report strongly experiencing MI principles was supported
  • This result supported the findings of previous research that MI is efficacious with offenders on the criterion of motivation to change (Clark, 2006; Czuchry et al., 2006; De Leon et al., 2000; Ginsburg et al., 2002; Harper & Hardy, 2000; Harry, 2005 ) The offenders in Anstiss’ (2005) study constituted low to medium risk offenders. It is therefore difficult to compare studies directly but together the two studies suggest that MI is efficacious with low, medium and high risk offenders. For some offenders, their motivation to change decreased from pre to post SMP. Process research might elucidate why. The finding that a brief intervention (e.g. SMP) is efficacious with high risk offenders partially contradicts the risk principle (Andrews & Bonta, 2006). However, this is based on recidivism not motivation. Further research measuring recidivism with high and low risk offenders, following MI, in the absence of other programmes, would clarify whether MI contradicts the risk principle. A survival analysis study is underway to measure recidivism with the offenders involved in this study.
  • Psychologists tended to show strengths in the use of global competencies, like those established in foundational clinical training, but rarely used the more specific, and perhaps sophisticated, skills of MI. Psychologists tended to only use simple reflections, similar to paraphrasing. While this demonstrated that they had understood, it did not allow a deeper understanding. Furthermore, reflections allow a clinician to reinforce change talk. When offenders used change talk, psychologists tended to move prematurely rather than reflect the statement and foster the offender’s thoughts about change. This has potential implications as studies have demonstrated that change talk, when fostered, leads to commitment language, which has predicted motivational outcomes (Amrhein et al., 2003).
  • To some degree the sample acted as their own control group because the same offenders were tested over time. Ideally a control group would have consisted of offenders randomly assigned or matched on key variables, such as risk of recidivism. Also, the process component ensured that the outcome data was not considered in isolation. This study relied partially on self-report measures, which when compared to behavioural observations can be vulnerable to cognitive biases. This was partially controlled for with the BIDR. The sample size for the one-way repeated measures ANOVA was expected to be too small to reach adequate statistical power and therefore alpha was increased. Cohen (1992) suggested that setting alpha above the traditional .05 level is defensible during exploratory research. Consequently, findings need to be considered in light of the increased risk of Type One error. Demand characteristics possibly affected the way psychologists acted when being audio-taped, and psychologists may not have passed on audio-tapes that they perceived as disparaging. To mitigate this, psychologists were reassured that their audio-tapes were confidential. The A5PMI did not undergo validity and reliability checks (other than internal consistency based on data used in this study) and instead was developed through a face validity approach and pilot testing.
  • K Austin The Efficacy Of Motivational Interviewing With Offenders

    1. 1. The Efficacy of Motivational Interviewing with Offenders The New Zealand Psychological Society Annual Conference 2009 Kevin Austin and Dr. Mei Wah Williams
    2. 2. Introduction <ul><li>I will: </li></ul><ul><ul><li>briefly outline relevant motivational interviewing (MI) and correctional rehabilitation literature </li></ul></ul><ul><ul><li>introduce and describe the current study </li></ul></ul><ul><ul><li>describe the results, discuss the findings and recommend further research </li></ul></ul>
    3. 3. Literature <ul><li>Motivation was historically (pre 1970s) viewed as a stable personality trait </li></ul><ul><li>In the late 1970s research demonstrated that therapist variables better predicted motivation </li></ul><ul><li>As such, attention shifted to how therapists could foster motivation </li></ul><ul><li>MI emerged as a “promising approach” </li></ul>
    4. 4. Literature <ul><li>Applications of MI proliferated </li></ul><ul><li>However, MI’s popularity out-stripped the evidence </li></ul><ul><li>Unlike MI, the efficacy of correctional rehabilitation was rigorously refuted (Martinson, 1974) </li></ul><ul><li>Scholars counter argued for the efficacy of correctional rehabilitation based on empirical studies </li></ul><ul><li>The “what works” approach emerged </li></ul>
    5. 5. Literature <ul><li>Risk, need and responsivity (RNR) </li></ul><ul><li>Motivation as a component of responsivity </li></ul><ul><li>MI was adapted for use with offenders </li></ul><ul><li>Like other fields, MI was used with offenders despite little empirical support </li></ul>
    6. 6. Research questions <ul><li>Can MI foster motivation to change among a high risk of recidivism offender sample? </li></ul><ul><li>How well is MI delivered to offenders? </li></ul>
    7. 7. Participants <ul><li>A power analysis to calculate sample sizes (26 for t-tests; 21 for ANOVA) </li></ul><ul><li>38 male offenders from Waikeria and Tongariro-Rangipo prisons (Roc*RoI M = .78, SD = 0.09) </li></ul><ul><li>18 - 42 years, M = 27.24, SD = 6.7 (younger than the general prison pop) </li></ul><ul><li>76.3% identifying as Māori (disproportionate number of Māori) </li></ul><ul><li>High criminal versatility </li></ul><ul><li>Fairly typical high risk offender profile (Wilson, 2004) </li></ul><ul><li>12 offenders consented to a follow-up assessment of motivation to change and programme integrity. </li></ul><ul><li>Five registered clinical psychologists </li></ul>
    8. 8. The Short Motivational Programme (SMP) <ul><ul><li>Delivered over five sessions by psychologists (manualised) </li></ul></ul><ul><ul><li>Focuses on rehabilitative needs </li></ul></ul><ul><ul><li>Reflects MI and correctional rehabilitation principles </li></ul></ul><ul><ul><li>High risk offenders – serving short sentences </li></ul></ul>Courtesy of: Dr J. Porter, PsyD Principles Skills Change Talk Commitment Talk Behaviour Change Spirit
    9. 9. Hypotheses <ul><li>Offenders will demonstrate a sustained increase in motivation to change. </li></ul><ul><li>Psychologists will demonstrate MI skills. </li></ul><ul><li>Offenders will experience MI principles. </li></ul>
    10. 10. Method <ul><li>Outcome: repeated measures - within group </li></ul><ul><li>Shift in motivation: pre- to post-intervention. </li></ul><ul><li>A sub-group consented to a third time period: follow-up </li></ul><ul><ul><li>University of Rhode Island Change Assessment Questionnaire – adapted for the SMP (SMP URICA) </li></ul></ul><ul><ul><li>Balanced inventory of desirable responding (BIDR) </li></ul></ul><ul><li>Process comprised two parts: </li></ul><ul><li>Psychologists’ use of MI skills </li></ul><ul><ul><li>Motivational interviewing treatment integrity code 3.0 (MITI) </li></ul></ul><ul><li>Offenders’ experiences of MI principles </li></ul><ul><ul><li>Assessment of the five principles of MI (A5PMI) </li></ul></ul>
    11. 11. Results - outcome <ul><li>A statistically significant increase in SMP URICA scores from pre-SMP to post-SMP for the total sample </li></ul>
    12. 12. Results - outcome <ul><li>For the sub-group, there was a significant and large effect for time (pre - post SMP - follow-up) </li></ul><ul><li>Pairwise comparison showed the significant effect was from pre - post-SMP; not from post-SMP to follow-up </li></ul>
    13. 13. Results - process <ul><li>Psychologists’ MI skills </li></ul><ul><li>Psychologists demonstrated full competency in using direction </li></ul><ul><li>Psychologists closely approached competence for global clinician ratings, percent open questions, evocation, collaboration, autonomy and empathy. </li></ul><ul><li>Psychologists did not achieve competency for reflection to question ratio, complex reflections and MI adherent behaviours (skills specific to MI). </li></ul><ul><li>The hypothesis that psychologists would demonstrate competence in the use of MI skills was not fully supported. </li></ul><ul><li>N = 5 </li></ul>
    14. 14. Results - process <ul><li>A score of 4 or above indicated an offender strongly experienced the MI principle </li></ul>
    15. 15. Discussion <ul><li>MI, a brief intervention, can effect a shift in motivation to change with high risk offenders </li></ul><ul><li>Somewhat contradicts the risk principle </li></ul><ul><li>Together, this study and Anstiss’ 2005 study, support the use of MI across the risk spectrum </li></ul>
    16. 16. Discussion <ul><ul><li>Offenders strongly experienced MI principles </li></ul></ul><ul><ul><li>MI principles tend to reflect general clinical competencies, such as expressing empathy </li></ul></ul><ul><ul><li>Psychologists’ strengths may reflect general clinical training rather than specific training in MI </li></ul></ul><ul><ul><li>The principles have better predicted shifts in motivation to change than MI skills </li></ul></ul><ul><ul><li>But, greater outcomes are likely if MI skills are also used, such as reinforcing change talk </li></ul></ul>
    17. 17. Limitations <ul><li>The omission of a control group </li></ul><ul><li>The use of self report measures </li></ul><ul><li>Increased risk of Type one error for follow-up group </li></ul><ul><li>Demand characteristics may have affected psychologists’ audio-taping </li></ul><ul><li>Difficult to generalise across risk profiles </li></ul>
    18. 18. Future studies <ul><li>Does the SMP alone (with high risk offenders) reduce re-offending and therefore contradict the risk principle </li></ul><ul><li>Norms and cut-off scores for the SMP URICA </li></ul><ul><li>How are the principles of MI and correctional rehabilitation effectively combined </li></ul><ul><li>There is a paucity of internal validity studies </li></ul><ul><li>Research on the use of MI with offenders remains in its infancy - a robust replication (addressing limitations) </li></ul>
    19. 19. Thanks to: <ul><li>The study participants (offenders and psychologists) </li></ul><ul><li>Dr. Mei Wah Williams; Massey </li></ul><ul><li>The Department of Corrections </li></ul><ul><ul><li>Glen Kilgour (Principal Psychologist) </li></ul></ul><ul><ul><li>Dr. Llewelyn Richards-Ward </li></ul></ul>