OARS Motivational Interviewing Skills: Examples and Practice

OARS is the set of four core communication skills in motivational interviewing: Open questions, Affirmations, Reflective listening, and Summaries (Miller & Rollnick, 2013). Many clinicians who have had a day of training can recite them. Fewer can do them under pressure, and self-assessment is a poor guide: in a randomized trial of 140 licensed substance-abuse professionals, clinicians’ self-ratings of their MI skill were unrelated to what trained coders heard on their tapes (Miller et al., 2004). The OARS motivational interviewing skills are easy to list and hard to perform, and the gap between the two is measurable.

This guide works through each OARS motivational interviewing skill with synthetic examples, shows how the MITI 4.2.1 fidelity code turns them into numbers, and ends with what the training literature says about practice that sticks.

What are the OARS skills in motivational interviewing?

OARS is the acronym Miller and Rollnick use for the four core communication skills of motivational interviewing: Open questions, Affirmations, Reflective listening, and Summaries (Miller & Rollnick, 2013). They are not a phase of MI. They are the micro-skills a clinician uses throughout all four MI processes, from engaging and focusing through evoking and planning.

Miller and Rollnick’s own definition puts the skills in their place: motivational interviewing is “a collaborative conversation style for strengthening a person’s own motivation and commitment to change” (Miller & Rollnick, 2013, p. 12). Two things about OARS follow from that and are easy to miss. First, the four OARS skills are not equally weighted. Reflective listening does most of the work, and fidelity instruments such as the MITI treat it that way. Second, OARS by themselves are not MI. A clinician can produce flawless open questions and reflections and still never steer toward the client’s own reasons for change. Miller and Rollnick (2013) frame the skills as serving the MI spirit of partnership, acceptance, compassion, and evocation. Without a direction toward change, the skills amount to good listening.

That is why fidelity measures such as the MITI 4.2.1 (the manual is free from the Motivational Interviewing Network of Trainers) count the OARS behaviors and also rate the whole segment for Cultivating Change Talk and Softening Sustain Talk. The counts tell you whether the skills were used; the globals tell you whether they went anywhere (Moyers et al., 2016).

O: how do you ask open questions in MI?

An open question in motivational interviewing is one the client cannot answer with a yes, a no, or a single fact, so it hands them the floor (Miller & Rollnick, 2013). The practical rule is simple: if the question could be answered with one word, rewrite it.

Closed: Do you want to cut down? Open: What would be different for you if you cut down?

Closed: Have you tried quitting before? Open: What happened the last time you tried?

Closed: Is your partner supportive? Open: How does your partner see all this?

One caution that trips up trainees who have been told “ask open questions.” A run of open questions is still a run of questions. MITI 4 merged open and closed questions into a single Question code and dropped the percent-open-questions benchmark that MITI 3.1 used (Moyers et al., 2016). In the current instrument, the question that matters is not “how open?” but “how many questions relative to reflections?” The answer to a good open question should usually be followed by a reflection, not another question.

A: what makes an affirmation land?

An affirmation lands when it names a specific strength, effort, or value drawn from the client’s own words (Miller & Rollnick, 2013); the MITI codes it as an MI-adherent behavior (Moyers, Manuel & Ernst, 2014). The word that matters is specific. “You’re doing great” is praise, and it comes from the clinician’s chair. An affirmation comes from the client’s own material.

Client: I didn’t drink on Tuesday. Big deal, it was one night.

Praise: That’s great, I’m really proud of you. Affirmation: You went a whole evening without it, on a night when you usually would. That took deciding, not just luck.

Two craft rules make affirmations credible. Start with “you,” not “I,” so the credit stays with the client. And affirm effort and values before outcomes, because outcomes can fail next week and the values are still there. The MITI codes the clinician’s Affirm behavior as one of three MI-adherent counts, alongside Seeking Collaboration and Emphasizing Autonomy (Moyers, Manuel & Ernst, 2014).

R: what is reflective listening, and why does the depth matter?

Reflective listening means making a statement, not asking a question, about what the client means (Miller & Rollnick, 2013). Depth matters because MITI 4.2.1 builds its two headline ratios from it: the reflection-to-question ratio and percent complex reflections (Moyers, Manuel & Ernst, 2014). The fidelity instruments split reflections in two. A simple reflection restates or rephrases what the client said and adds little or nothing. A complex reflection adds meaning or emphasis beyond the client’s words: an inferred feeling, a metaphor, the other side of their ambivalence, or a guess at what comes next (Moyers, Manuel & Ernst, 2014).

The same client line can be met with four moves of increasing depth. Pick a scenario and climb the ladder.

The reflection ladder: one client line, four clinician moves

Drinking after a layoffA daughter and the pillsWalking, after marathons

Client
I know the drinking's gotten heavier since the layoff, but it's the only thing that shuts my head off at night.

1. 1Closed questionQ
2. 2Open questionQ
3. 3Simple reflectionSR
4. 4Complex reflectionCR

Clinician

“Sleep has become the thing you're protecting, and the drinking is the price you're paying for it.”

MITI code: Complex reflectionadds a reflection and raises % complex reflections

Adds meaning the client did not say: the trade-off, and the cost side of the ambivalence. This is the move the MITI rewards most.

Synthetic exchanges. Codes follow MITI 4.2.1 as paraphrased in our MI practice tool; the ladder is a teaching device, not a rule that deeper is always better.

Move	Example (client: "the drinking is the only thing that shuts my head off at night")	MITI 4.2.1 code	Effect on the ratios
Closed question	Do you drink every night?	Question	Adds a question; R:Q falls
Open question	What does shutting your head off do for you?	Question	Adds a question; R:Q falls
Simple reflection	The drinking is the one thing that quiets your mind at night.	Simple Reflection	Adds a reflection; R:Q rises, %CR falls
Complex reflection	Sleep has become the thing you're protecting, and the drinking is the price you pay for it.	Complex Reflection	Adds a reflection; R:Q and %CR both rise

Why does the depth matter so much? Because the MITI’s two headline ratios are built from it. The reflection-to-question ratio (R:Q) divides all reflections by all questions, with 1:1 suggested as fair practice and 2:1 as good. Percent complex reflections divides complex reflections by all reflections, with 40% fair and 50% good (Moyers, Manuel & Ernst, 2014). The manual states that these thresholds are expert opinion without normative data, so read them as a direction of travel rather than a cutoff.

Miller and Rollnick (2013) describe several ways a reflection can add meaning, and each is a different route to a complex reflection. The stems below are the ones trainers hand out; the point is not the wording but the fact that each one is a statement.

Skill	Stems that work	What it adds
Open question	What would be different if…? / How does that fit with…? / Tell me about the last time…	Hands the client the floor; evokes their reasons, not yours
Affirmation	You decided to… / It took something to… / What stands out is how much you value…	Names a specific strength or effort in the client's own material
Simple reflection	You're… / So the drinking… / It sounds like…	Restates; shows you heard it
Reflection of feeling	Part of you feels… / That left you…	Names the emotion underneath the words
Double-sided reflection	On the one hand… and on the other…	Holds both sides of the ambivalence; ends on the change side
Amplified reflection	There's no reason at all to change anything here.	Slightly overstates the sustain talk so the client argues the other side (use with care)
Continuing the paragraph	…and that's why you're starting to wonder.	Says the next sentence the client was about to say
Summary	Let me see if I have this… What have I missed?	Collects, links, and transitions; ends by returning the floor

One coding rule shapes how reflections should be delivered. The MITI codes at most one reflection per clinician volley, and if any part of the volley is complex, the whole volley counts as one Complex Reflection (Moyers, Manuel & Ernst, 2014). Stacking three reflections in one breath earns the same count as one. The skill is to reflect once, stop, and let the client respond.

S: when do you summarize, and what goes in?

Summarize at natural transition points: the end of a topic, after a run of change talk, or before moving to planning. A summary in MI is a longer reflection that collects several things the client has said and hands them back as a whole (Miller & Rollnick, 2013). Miller and Rollnick distinguish three jobs a summary can do: collecting (gathering change talk as it accumulates), linking (connecting what the client said now with something said earlier), and transitional (closing one part of the conversation and moving to the next).

Clinician: Let me see if I’ve got this. The drinking started as a way to sleep after the layoff, and it still does that job. At the same time you’ve noticed the mornings are worse, your daughter has stopped calling, and you said the word “trapped” twice. You’re not sure what you want to do about it yet. What have I missed?

Notice the construction. The summary leads with the client’s own change talk, includes the sustain talk without arguing with it, ends with an open question that returns the floor, and does not add the clinician’s opinion. In the MITI a summary is coded as a reflection, simple or complex depending on what it adds, so it also feeds the R:Q ratio (Moyers, Manuel & Ernst, 2014).

How do OARS skills show up in a MITI score?

OARS skills become numbers through behavior counts: a trained coder listens to a 20-minute segment and tallies each clinician utterance as a Question, Simple Reflection, Complex Reflection, Affirm, or one of the other MITI 4.2.1 codes (Moyers, Manuel & Ernst, 2014). Here is a synthetic five-volley exchange, coded.

Client: I’m here because my GP said I had to be. The drinking isn’t the problem, the job is.

Clinician: You’d rather be anywhere else, and from where you sit the drinking is a symptom, not the illness. [Complex Reflection]

Client: Exactly. Fix the job and the rest sorts itself out.

Clinician: What would “fixed” look like? [Question]

Client: Something that doesn’t leave me wired at eleven at night, needing three beers to come down.

Clinician: Three beers to come down. [Simple Reflection]

Client: Some nights more. It’s not every night.

Clinician: You’re keeping an eye on it even while you’re telling me it isn’t the problem. [Complex Reflection]

Client: I suppose I am.

Clinician: Would it be useful to look at the nights it’s more than three, or would you rather stay with the job for now? [Seeking Collaboration]

Three reflections against one question gives an R:Q of 3:1, and two of those three reflections are complex, so %CR is 67%. Both clear the manual’s “good” line. The last turn is coded Seeking Collaboration, an MI-adherent behavior, rather than a Question, because its job is to share the agenda (Moyers, Manuel & Ernst, 2014). A reader can check every one of those decisions against the text, which is the point of coding utterances rather than impressions. For the full set of counts and the four global ratings, see our guide to MITI motivational interviewing coding.

What the counts miss is the client side. The MITI codes the clinician only; whether all those reflections produced change talk is a question for the MISC change-talk codes, and the two instruments are compared in MISC vs MITI.

Why do OARS skills matter for outcomes?

OARS matter because the way a clinician talks measurably shapes what the client says next, and part of what the client says predicts outcome. In a meta-analysis of 12 studies of MI’s causal chain, clinician MI-consistent skills correlated with more client change talk (r = .26), and client sustain talk predicted worse outcomes (r = −.24); change talk itself was not significantly related to outcome in that pooled analysis (Magill et al., 2014). MI itself produces a small but reliable benefit, about g = 0.28 against weak or no-treatment comparisons across 25 years of studies (Lundahl et al., 2010).

Magill and colleagues’ (2014) meta-analysis also found that MI-inconsistent clinician behavior (confronting, persuading without permission) was associated with more sustain talk (Magill et al., 2014). One reading is that OARS are protective partly because they are what a clinician does instead of arguing. Every reflection is a turn in which the clinician did not make the case for change, and so did not invite the client to make the case against it. The parallel in medical consultations is well documented in the Empathic Communication Coding System, where the clinician’s response to an emotional cue is coded along a similar ladder from ignoring to exploring.

How do you practice OARS so the skills stick?

Practice OARS by recording role-plays, coding them, and getting feedback on the coding; three to four feedback or coaching sessions over six months are enough to sustain the skills (Schwalbe, Oh & Zweben, 2014). A workshop alone produces real gains; the question is what happens afterwards. In an early randomized trial of MI training methods, 140 licensed clinicians were assigned to one of five arms: a workshop alone, workshop plus feedback, workshop plus coaching, workshop plus both, or a waiting-list control given self-guided training materials. Coaching and feedback raised post-training proficiency beyond the workshop, proficiency was generally maintained over 12 months of follow-up, and every gain was measured from audiotaped practice rather than from what clinicians said about themselves (Miller et al., 2004).

Schwalbe, Oh and Zweben’s (2014) meta-analysis of 21 MI training studies quantified how skills decay after a workshop.

Effect sizes from Schwalbe, Oh and Zweben’s (2014) meta-analysis of MI training studies. Training raises skill (d = 0.76); over the following six months skills erode without feedback or coaching (d = −0.30) and hold with it (d = 0.03). Studies with low trainee attrition even showed a small further gain (d = 0.12).

The authors’ practical conclusion was that three to four feedback or coaching sessions over six months are enough to sustain skills (Schwalbe, Oh & Zweben, 2014). Two caveats: the trainees were mostly agency and substance-use treatment staff, and the effect sizes pool pre-post change within studies rather than randomized contrasts, so they describe the pattern rather than prove the mechanism. Feedback here means someone coded a recording and told the clinician what their R:Q ratio and %CR actually were. That is a very different thing from a trainee’s sense of how it went, which Miller and colleagues (2004) found to be uncorrelated with the tapes.

Here is a practice loop that matches what the evidence rewards:

1. Record a role-play of ten to twenty minutes with a colleague playing an ambivalent client. The trials above measured skill from recordings, so practice the way it is measured.
2. Code every clinician turn with the MITI behavior codes: Question, Simple Reflection, Complex Reflection, Affirm, Seeking Collaboration, and the rest.
3. Compute R:Q and %CR, and look at where the questions cluster. A run of questions is the usual way R:Q falls, and it tends to happen right after the client says something hard.
4. Rewrite the weakest three turns as complex reflections. Say them out loud.
5. Repeat monthly for six months, ideally with a second coder on at least one tape so you can check your own coding.

Our free motivational interviewing practice tool is built for steps two to four. Paste the transcript, tag each turn with a keystroke, and the ratios update as you go. Two drills train the ear separately: name the skill in a given clinician line, and pick the reflection that goes deeper. Nothing leaves the browser (see our privacy policy).

What are the most common OARS motivational interviewing mistakes?

The most common OARS errors are asking questions in runs, disguising closed questions as reflections, stacking reflections, praising instead of affirming, and summarizing only the sustain talk. Most are invisible to the clinician making them:

• Asking in runs. Three open questions in a row feel like curiosity and code as three Questions. Follow a question with a reflection before asking again.
• Reflecting with a rising inflection. “So you’re feeling trapped?” said as a question invites a yes or no and hands the floor back to the clinician. Miller and Rollnick (2013) recommend turning the voice down at the end; a reflection is a statement.
• Stacking reflections. Several reflections in one breath count once under MITI’s one-per-volley rule (Moyers, Manuel & Ernst, 2014). Reflect, stop, listen.
• Praising instead of affirming. “I’m proud of you” credits the clinician. “You decided to go, and you went” credits the client.
• Summarizing only the sustain talk. A summary that faithfully repeats every reason not to change is accurate and counterproductive. Lead with the change talk the client actually offered.
• Treating the ratios as a scoreboard. R:Q 2:1 and %CR 50% are expert opinion, not validated cutoffs, and a session can hit both with shallow, aimless reflections (Moyers, Manuel & Ernst, 2014). The MITI globals exist to catch that.
• Trusting your own sense of how it went. It does not track the tape (Miller et al., 2004). Code a recording.

If you change one thing, change this: reflect twice for every question you ask, and make half of those reflections say something the client did not. In the training studies above, OARS motivational interviewing skills gained more when trainees got feedback or coaching on recorded practice (Miller et al., 2004), and eroded within six months when they did not (Schwalbe, Oh & Zweben, 2014).

References

• Miller, W. R., & Rollnick, S. (2013). Motivational Interviewing: Helping People Change (3rd ed.). New York: Guilford Press.
• Moyers, T. B., Manuel, J. K., & Ernst, D. (2014). Motivational Interviewing Treatment Integrity Coding Manual 4.2.1. University of New Mexico, Center on Alcoholism, Substance Abuse, and Addictions (CASAA). Manual PDF via the Motivational Interviewing Network of Trainers
• Moyers, T. B., Rowell, L. N., Manuel, J. K., Ernst, D., & Houck, J. M. (2016). The Motivational Interviewing Treatment Integrity Code (MITI 4): Rationale, preliminary reliability and validity. Journal of Substance Abuse Treatment, 65, 36–42. doi:10.1016/j.jsat.2016.01.001
• Miller, W. R., Yahne, C. E., Moyers, T. B., Martinez, J., & Pirritano, M. (2004). A randomized trial of methods to help clinicians learn motivational interviewing. Journal of Consulting and Clinical Psychology, 72(6), 1050–1062. doi:10.1037/0022-006X.72.6.1050
• Schwalbe, C. S., Oh, H. Y., & Zweben, A. (2014). Sustaining motivational interviewing: a meta-analysis of training studies. Addiction, 109(8), 1287–1294. doi:10.1111/add.12558
• Magill, M., Gaume, J., Apodaca, T. R., Walthers, J., Mastroleo, N. R., Borsari, B., & Longabaugh, R. (2014). The technical hypothesis of motivational interviewing: A meta-analysis of MI’s key causal model. Journal of Consulting and Clinical Psychology, 82(6), 973–983. doi:10.1037/a0036833
• Lundahl, B. W., Kunz, C., Brownell, C., Tollefson, D., & Burke, B. L. (2010). A meta-analysis of motivational interviewing: Twenty-five years of empirical studies. Research on Social Work Practice, 20(2), 137–160. doi:10.1177/1049731509347850

If you run MI training or supervision and need the coding done at scale, Tagaroo turns the MITI behavior counts into an evidence-anchored annotation workflow, with an AI first pass your coders review and inter-rater reliability computed as they work. See how that fits a training program end to end in Tagaroo for supervision and training, and for the client side of the conversation, the MISC change talk and sustain talk codes.