Online Training That Changes Behaviour, Not Just Confidence: What the Data Says to Do Instead
Most online training still leans on slides, videos, and content-heavy modules that boost confidence but barely touch real-world behaviour. Research on online feedback training shows that role-play-based practice outperforms passive formats, especially when communication skills are involved.
In the first part of this series, we looked at a piece of research comparing three types of online training for feedback skills: a slideshow, a video, and a live online practice session with a role-play coach.
The punchline was simple:
Slides and video made people feel more confident.
Only the practice-based online training noticeably improved what they actually did in a real conversation.
So if you’re responsible for online training in your organisation, the real question should be “How do we design online training that changes behaviour, not just confidence?” (Courtesy photos from Freepik)
Authentic learning vs passive online training
The study on feedback sits on top of a wider body of research about authentic learning that focuses on real-world, socially rich scenarios where people actively solve complex problems – versus more passive forms such as slideware, lectures, and instructional videos.
A few patterns keep showing up:
Authentic learning places people in realistic environments, with roles and responsibilities that mirror the workplace, and asks them to act, not just listen.
Role play – in the sense of imagining you are yourself or someone else in a specific situation and acting as you genuinely would – is a particularly powerful tool for communication skills, because it forces you to choose words, manage your emotions, and respond in real time.
Practical, exploratory learning of this kind consistently outperforms theory-only approaches when you care about transfer into real-world performance.
Despite that, many organisations still default to online training that is:
Heavy on presentation and explanation
Light on practice and feedback
But after the shift to video-conferencing, the logistics argument is weaker than it used to be.
“The research argues that running role-play-based online sessions is now only “marginally more difficult” than producing and delivering slides or recorded videos, while delivering better behavioural outcomes.”
Three traps that keep online training stuck
Putting the evidence together, you can see at least three common traps.
1. The confidence trap
We ask people if they feel more confident after online training. They say yes. We conclude the programme “worked”.
But in the feedback study, self-reported confidence went up in every group – slides, video, and practice – while the actual behavioural gains were heavily concentrated in the practice group. The correlation between “I feel confident” and “I gave effective feedback” was essentially zero.
The researchers suggest that self-reported confidence is easily influenced by expectations (“I’m supposed to feel more confident now”) and by participants wanting to be seen as responsive to the training. (Courtesy photos from Freepik)
That doesn’t mean confidence is irrelevant. It just means:
If “people say they feel more confident” is your main success metric for online training, you’re measuring the wrong thing.
2. The content trap
There’s a lot of energy in L&D around upgrading content:
Turning slide decks into videos
Turning videos into interactive e-learning
Re-branding the LMS as an “academy”
In the study, changing content format – from slideshow to video – didn’t make a meaningful difference to behaviour. Only changing the nature of the intervention, from content to practice, did.
“If most of your online training investment is going into reshaping the same content in different passive formats, you’re optimising the packaging, not the outcome.”
3. The logistics myth
Role play and simulation are often dismissed as “too hard to organise”, especially online.
Yet the paper points out that the widespread adoption of video-conferencing has removed many of the historic barriers (rooms, travel, scheduling blocks). Participants in the study experienced the online simulations as a realistic mirror of how their own workplaces had changed during COVID-19.
In short:
The idea that online role play is impractical is now more myth than fact.
The harder part is having the will to prioritise it.
Designing online training for behaviour change
So what do you actually do with this, if you’re designing or commissioning online training?
1. Start with the moment of use
Instead of starting with “what content should we cover?”, start with:
“What are the moments that matter in this skill?”
“What would we see and hear if someone was doing this well?”
For feedback, the study anchored everything around one moment:
“Can this person give clear, constructive feedback to a difficult colleague in a live conversation?”
Your equivalent might be:
Calling out inappropriate behaviour in a client meeting
Saying no to an unrealistic deadline
Running a psychologically safe team debrief after a mistake
Effective online training should simulate those moments and let people rehearse them, with feedback, not just describe them in theory.
2. Make practice the non-negotiable
When time gets tight, practice is often the first thing to be cut:
“We’ll still cover the model, but we won’t have time for role play.”
The evidence suggests that’s exactly backwards. Practical exploration via role play yielded better results than theory-only approaches and is particularly suited to communication skills like feedback. (Courtesy photos from Freepik)
When you’re planning online training, a useful design question is:
“How much of the live time is spent letting people actually try things out?”
If the answer is “five minutes at the end, if we get there”, it’s not really practice-based.
3. Build in proper debriefs
“Practice without reflection is just repetition. Practice plus reflection becomes learning.”
The study notes that while the intervention was effective, one limitation was the depth of debrief – the space for participants to make sense of what happened and link it back to their real work.
For your online training, that might look like:
Short, structured debrief questions after simulations
Encouraging people to name what they tried, what they noticed, and what they’d do next time
Inviting them to connect the exercise to specific relationships or meetings on their calendar
4. Redefine what “good” looks like
If behaviour change is the goal, then “good online training” can’t just mean:
High completion
High satisfaction
High self-reported confidence
More meaningful indicators might include:
Managers noticing and describing different behaviours in real meetings
Participants reporting specific conversations they handled differently after the programme
Observable changes in how feedback, challenge, or dissent show up in the team
The feedback study’s use of external assessors rating real conversations is a reminder: when we watch what people actually do, the picture gets a lot clearer.
The future of online training: beyond more content
None of this is an argument against online training. In fact, the research is quietly optimistic:
The entire intervention – from intro to coaching to simulation – was delivered fully online, via everyday devices.
Participants saw it as a realistic reflection of how their workplaces had shifted and as a model for future online learning.
The author concludes that role-play-based online training is a relatively underused but highly promising methodology for organisations that care about communication competence.
The real shift is from passive online training to practice-based online training – from counting modules completed to caring about what shows up in the next difficult conversation.
If your training looks good on paper but not on Monday…
If your online training looks great on paper but doesn’t quite show up in the moments that matter, we’d be happy to compare notes.
