In summary:
- Simulations can act as a bridge for students during the transition from academic theory to high-stakes real-world practice.
- These tools provide ‘training wheels’ where learners can build competence and fail safely without risking public safety or damaging their professional confidence.
- While psychology previously lagged due to its ‘nuanced, relational’ nature, generative AI now enables the dynamic and personalised interactions required for effective training.
- ‘Shadow box’ feedback pauses scenarios at key decision points, requiring learners to explain their reasoning rather than receiving simple right or wrong feedback.
- Dr Morrison advocates for closed loop systems – where AI only draws from vetted textbooks and research – to mitigate algorithmic bias and protect client confidentiality.
- This research is featured in the Australian Psychologist under the new category, Teaching and Training Initiatives and Innovations, designed to encourage new learning approaches in the field.
- Information about submission requirements is available on the journal's website. Submit to Australian Psychologist
The transition from university into real-world psychology settings, be that in a school, in a workplace or in a clinic, can often be a jarring moment for an early career psychologist.
After years of mastering knowledge at a theoretical level – postgraduate students are often then thrust into real-world placements where they must navigate the messy, ambiguous and high-stakes nature of human interaction.
This can feel like an ‘overnight shift’ from theory to practice, says Dr Ben Morrison, Associate Professor and organisational psychologist.
While there are clear safeguards in place to manage any potential risk that could arise during the transition from university into supervised practice, Dr Morrison, who is also an Associate Editor of the Australian Journal of Psychology, says any further opportunities to practice the critical skills required to make this transition should be embraced.
In a recent paper that Dr Morrison co-authored for Australian Psychologist, he and his co-authors explore the potential for virtual client simulations to serve as a complementary tool in this professional journey.
By leveraging generative AI and interactive technology, he says these simulations offer a "safe space" for both students and seasoned practitioners to rehearse complex conversations and refine critical skills.
The evolution of the virtual client
While the concept of virtual patients has existed for some time in medicine and nursing, the application of virtual clients in psychology has been more difficult to achieve, according to Dr Morrison.
A scoping review conducted by Dr Morrison and his co-researchers found only nine papers evaluating this type of training specifically within psychology.
"Psychology relies heavily on nuanced, relational, conversational, interactive elements, which traditionally have been harder to simulate," he says.
Early iterations were often described as rigid and procedural. However, the proliferation of generative AI has changed the landscape, allowing for more dynamic, personalised interactions.
"We're really only seeing the start of the full potential for psychology [now] because of the advances in interactive technology," he says.
For students, virtual simulations act as additional "training wheels", he says.
The paper notes that supervised practice placements offer a critical foundation to "ensure that students gain comprehensive exposure to a diverse range of clients and professional competencies".
Dr Morrison sees virtual client simulations as a supportive add-on to complement this important part of a psychology student’s learning journey, and offers further practice in managing “real-world, real-time, ambiguous situations”,
Virtual clients can support psychology students to:
- Rehearse conversations: "Students can have those rehearsed conversations and can practice making decisions and making judgments on the spot."
- Navigate complexity and experience variability: Being able to rehearse complex scenarios that they might not be exposed to in placements in a low-stake environment.
- Build competence: "Students can take [conversations] out of that high-stakes setting and have a safe space where they can start to build competence."
- Fail safely: Protect students from "taking a hit to confidence because they're making errors" that have real-world consequences.
Dr Morrison suggests a scaffolded approach where simulations begin with fundamental skills and "dial up the complexity" as the student progresses, eventually incorporating a diverse range of client personas and scenarios.
‘Shadow box’ training
A critical component of these simulations is the feedback mechanism. Dr Morrison’s research utilises ‘shadow box’ training, a concept developed by Gary Klein.
"[This is when] you're presenting the learners with realistic scenarios, and then at key decision points... you're asking them to make decisions or judgments, and we then pause the simulation essentially and provide feedback on their responses."
Unlike surface-level feedback (right vs. wrong), shadow boxing walks through the reasoning behind a choice.
With the integration of generative AI, this has evolved into a conversational experience. The AI can act as a coach, asking the user to reflect on their performance, which Dr Morrison notes is "really good for building self-reflection and critical thinking skills".
Beyond university students
The utility of virtual clients is not restricted to psychology training. For seasoned psychologists, these tools offer a robust platform for continuous professional development (CPD).
"Our training windows are relatively small compared to our careers, and so it's not really reasonable to expect practitioners to have seen everything," says Dr Morrison.
As practitioners navigate increasingly complex client presentations, virtual simulations offer a safe environment to parse through the 'noisiness' of real-world practice.
By exploring ethical tensions and competing priorities at a surface level, these tools can help psychologists refine the critical judgment necessary to make sense of ambiguous scenarios.
Some uses for seasoned psychologists include:
- Expanding exposure: "Psychologists might have only seen a limited number of client presentations, scenarios and so these simulations can help expand exposure."
- Practice new techniques: "Just like the students starting out, psychologists can rehearse, they can look at rare cases and more complex situations. They can also practice new techniques."
Dr Morrison can also see many different use cases across endorsement areas. For example:
Sports psychology: "You could imagine sports psychologists having clients who are experiencing performance anxiety before a tournament, and so they could practice conversing with a client with this type of presentation."
Forensic psychology: "Forensic psychologists [could explore] what it's like to have a client that's going to be quite resistant, that has other contextual factors to consider, like risk assessment in a forensic environment."
Organisational psychology: "For us in the organisational psychology team, we were looking at examples like conducting job analysis specific interview techniques."
Clinical psychology: "From a more clinical perspective, you could simulate psychological assessment and feedback sessions, where trainees practice communicating results, managing client reactions and maintaining a therapeutic alliance.”
Navigating ethics, bias and confidentiality
As with any AI-driven technology, virtual clients come with significant ethical considerations.
Dr Morrison identifies algorithmic bias and data confidentiality as the two primary hurdles.
"We need to avoid creating simulations that are 'garbage in, garbage out', or that reinforce stereotypes or oversimplify cultures and cultural contexts," he says.
To mitigate these risks, Dr Morrison advocates for closed-loop systems. In these systems, the AI draws only from a controlled "universe of materials", such as reliable journals, texts, papers that evidence-based.
While the path forward involves navigating ethical tensions and competing priorities, the potential for virtual client simulations to improve the quality of psychological practice is clear.
As Dr Morrison notes, "We can't bury our heads in the sand to these types of technologies.
This is one way to support the next generation of psychologists to be future-ready.
A common barrier to adoption is the fear that these tools are intended to replace human interaction or provide shortcuts in training. Dr Morrison is emphatic that this is a complementary tool.
"I definitely think that psychology is fundamentally human-centered," he says.
He believes this technology can "enhance our humanness" by allowing us to isolate and practice specific skills such as active listening, rapport building, and establishing a therapeutic alliance in a controlled setting.
"This technology can help us be better practitioners by reflecting on our responses."
The research paper this article is referencing was published in the new Australian Psychologist article category of Teaching and Training in Initiatives and Innovations.
If you have a psychology teaching and training initiative or innovation, you're welcome to submit an article in this category, which is designed to stimulate reflective practice about psychology teaching and training, and encourage the development and application of new engaging, authentic learning approaches with relevance to professional practice and public policy. Information about the requirements is available in the journal. Submit to Australian Psychologist.