TherapyCoach

About Features Evidence Sign In Launch

Why simulated practice matters

The research behind TherapyCoach — and the training gap it was built to close.

Therapists don't automatically improve with experience

One of the most uncomfortable findings in psychotherapy research is that therapists don't reliably get better over time. A landmark study tracking 6,591 patients across 170 therapists found that clinical outcomes actually declined slightly with increasing years of experience. More clients seen, more years practised — and outcomes got worse, not better.

Therapists, on average, did not improve over time — their patient outcomes actually worsened slightly as they gained more experience.

Goldberg et al. (2016), Journal of Counseling Psychology

This isn't because clinicians stop caring. It's because passive experience — seeing clients, attending workshops, reading — doesn't produce the kind of learning that translates into measurable skill improvement. Something else is needed.

Workshops and lectures don't change practice

Most continuing professional development for therapists still relies on formats that research has consistently shown to be insufficient. One-off workshops, conference presentations, and didactic training may increase knowledge, but they rarely change what a clinician actually does in the room with a client.

A systematic review of CBT training approaches found that instructor-led and self-guided training produced only moderate increases in competence — and the gains were strongest when training targeted specific, highly structured skills rather than general theory. Broader reviews of therapist training have found that without structured follow-up, ongoing practice, and feedback, the effects of workshop-based training wash out within months.

Standard training formats lack five critical processes: observing your own work, receiving expert feedback, setting incremental goals, repetitive behavioural rehearsal, and continuous performance assessment.

Rousmaniere (2017), Deliberate Practice for Psychotherapists

Deliberate practice closes the gap

The concept of deliberate practice — originally developed by Ericsson to explain expert performance in fields like music and sport — has been increasingly applied to psychotherapy training. Unlike routine practice, deliberate practice involves focused, effortful rehearsal of specific skills with immediate feedback, pushing just beyond current ability.

When applied to clinical training, deliberate practice means repeatedly rehearsing therapeutic micro-skills (reflections, open questions, validation sequences, Socratic dialogue) in a structured way, with feedback on performance — not just thinking about them or watching someone else do them.

Early evidence for deliberate practice in psychotherapy is promising. Trainees who engaged in structured skill rehearsal with feedback showed faster skill development than those using traditional training alone. The challenge has always been access: deliberate practice requires a practice partner, a realistic scenario, and timely feedback — resources that are expensive, logistically difficult, and rarely available outside of formal training programmes.

Simulation fills the access gap

Simulation-based training has a long evidence base in medicine, nursing, and allied health. A meta-analysis of 50 studies on virtual patient simulations found large effect sizes on both knowledge and skill outcomes. In social work, 61% of students achieved mastery in targeted engagement and assessment skills through simulated practice.

More recently, research has explored AI-driven client simulation specifically for counselling skills. Standardised patient roleplays — whether human or AI-generated — are consistently rated as highly valuable learning experiences that build practitioner confidence and allow safe experimentation with technique. The combination of realistic practice with structured feedback produces better skill transfer than either element alone.

Roleplay simulations with feedback were perceived as highly valuable opportunities that build clinical confidence and enable safe skill experimentation.

Räuchle et al. (2025); Maurya (2024)

The emergence of large language models now makes it possible to create client simulations that respond naturally and unpredictably — more closely approximating the experience of real human-to-human roleplay than the branching-logic simulations of the past.

What TherapyCoach does with this evidence

TherapyCoach was designed to bring together the components that research identifies as necessary for effective skill development: realistic practice scenarios, modality-specific coaching feedback, validated assessment scoring, and structured progression through skills of increasing difficulty.

It doesn't replace supervision. It fills the space between learning about therapy and doing therapy — the same space that flight simulators fill for pilots, and that simulation labs fill for surgeons. A space where you can practise, make mistakes, get feedback, and try again — without any risk to a real person.

Clinical fidelity frameworks

Coaching feedback and assessment scoring in TherapyCoach are calibrated against established clinical fidelity instruments — the same tools used in research and supervision to evaluate therapeutic competence.

Motivational Interviewing — feedback is informed by the Motivational Interviewing Treatment Integrity (MITI 4.2.1) coding manual, which evaluates global dimensions (cultivating change talk, softening sustain talk, partnership, empathy) and behavioural counts (reflections, questions, affirmations, seeking collaboration, persuade, confront). The MITI is the most widely used MI fidelity measure in research and clinical training.

Cognitive Behavioural Therapy — feedback is informed by the Revised Cognitive Therapy Scale (CTS-R), which assesses therapist competence across domains including agenda setting, guided discovery, conceptual integration, application of CBT methods, and homework setting. The CTS-R is used internationally as a training and supervision benchmark.

Dialectical Behaviour Therapy — feedback is informed by the DBT Adherence Coding Scale, which evaluates adherence to DBT strategies including dialectical strategies, core mindfulness, distress tolerance, emotion regulation, and interpersonal effectiveness, as well as structural elements like diary card review and behavioural chain analysis.

These instruments are used as reference frameworks for structuring feedback — TherapyCoach does not claim to produce validated fidelity scores. AI-generated scores are approximations intended to support learning, not to replace formal competency assessment or clinical supervision. For more on how TherapyCoach uses AI responsibly and aligns with Australian regulatory frameworks, see our AI & Compliance page.

Ready to practise?

Try a session with a simulated client in MI, CBT, or DBT.

Launch TherapyCoach

References

Clinical fidelity instruments