In branching scenarios, we can use a combination of immediate and delayed consequences and feedback.
Consequences are what happens as a result of decisions: what they see, hear, or experience.
Feedback is what we tell learners after decisions.
Both consequences and feedback can be either immediate or delayed. We can use these approaches together to provide learners with information to learn from their mistakes and successes. You may not use all of these in every scenario (in fact, you probably won’t), but this flowchart shows when each one would happen. I’ll break down each component separately.
Use immediate consequences often
Immediate consequences are the intrinsic effects of decisions. A customer who responds angrily, software that doesn’t produce the desired result, or time lost on a project could all be immediate consequences. These consequences don’t directly tell the learner, “Sorry, that was incorrect.” Learners have to perceive and understand the cues in the scenario. They have to draw conclusions based on those cues.
If your learners will need to follow cues when they apply what they’re learning, it’s helpful to provide real-world consequences in your scenario. It’s beneficial to practice interpreting cues.
Immediate consequences that simulate real-world cues can also be more engaging than the omniscient narrator dictating what you did right or wrong. It keeps learners in the mindset of the story without hitting them over the head with a reminder that they’re learning something.
Use immediate feedback with novices
Immediate feedback is different from intrinsic consequences. This is the instructional feedback or coaching that directly tells learners why their decisions are right or wrong. While this can pull people out of the “flow” of a story, immediate feedback can be helpful in some situations.
First, novice learners who are still building mental models of a topic may benefit more from immediate feedback. Novices may not have the expertise to sort through real-world cues and draw accurate conclusions from them.
Therefore, it may be more important to provide immediate feedback after each decision in a branching scenario if your audience is new to the topic. In his research report “Providing Learners with Feedback (Part 1),” Will Thalheimer explains the benefits of immediate feedback for novices.
“On the surface of it, it just doesn’t make sense that when a learner is piecing together arrays of building blocks into a fully-formed complex concept, they wouldn’t need some sort of feedback as they build up from prerequisite concepts. If the conceptual foundation they build for themselves is wrong, adding to that faulty foundation is problematic. Feedback provided before these prerequisite mental modelettes are built should keep learners from flailing around too much. For this reason, I will tentatively recommend immediate feedback as learners build understanding.”
Will Thalheimer
Use delayed consequences when realistic
We don’t always immediately know our mistakes are wrong in real life. Sometimes the consequence isn’t obvious right away. Sometimes it seems like a gain the short run, but causes problems in the long run. If that’s the kind of situation you’re training for, letting people continue on the wrong path for a little while makes sense. Neither limited branching nor immediate failure allow you to show delayed consequences.
Providing these delayed consequences has the advantage of better learning from delayed feedback, plus it creates a more realistic and engaging story. Delayed consequences shouldn’t be forced into a scenario where it’s not realistic, but they are a good way to show the long-term effects of actions.
Examples of delayed consequences
Think about how delayed consequences could be shown in these examples:
Free drinks: A bartender gives away many free drinks. The immediate consequence is that the customers are happy, but the delayed consequence is a loss of profit for the bar.Less expensive product sold: A sales associate sells a customer a product that is less expensive but meets the customer’s needs. The immediate consequence is that the sales associate makes less commission that day, but the delayed consequence is that the customer is loyal and refers 2 friends. In this case, the total commission earned is higher even though the immediate sale was lower.Missed question: A doctor could skip a screening question with a patient. The immediate consequence is finding something that looks like the problem based on incomplete information, but the delayed consequence is the actual underlying problem remaining.Training request: A manager asks an ID to create training. The ID gets started building it right away, trusting that the team requesting the training knows their needs. The immediate consequence is a happy manager, but the delayed consequence is ineffective training that doesn’t actually solve the business problem.Small ethical lapse: If you’re teaching ethics, a small ethical lapse early in the scenario might not seem like a big deal. The immediate consequence might be meeting a deadline or increased recognition. In the long run, that small lapse leads a continued need to cover up your actions. The Lab: Avoiding Research Misconduct is an example with delayed consequences in some paths.
Use delayed feedback with experienced learners
Thalheimer notes that delayed feedback may be more effective for retention (i.e., how much do learners remember). That effect might be due to the spacing effect (that is, reviewing content multiple times, spaced out over times, is better for learning than cramming everything into a single event). The delay for feedback doesn’t have to be long; one study mentioned in Thalheimer’s report showed that delaying feedback by 10 seconds improved outcomes.
Megan Sumeracki of the Learning Scientists blog also notes the connection between delaying feedback and the spacing effect. (The experiments she describe in this article use much longer timelines for feedback than what I’m talking about with scenarios, but I still find that the quote resonates.)
Logically if we think about feedback as correcting errors, then it makes sense that we would want pretty immediate feedback. But if we think about feedback as another presentation of the information, then a space ought to improve learning.
–Megan Sumeracki
Delayed feedback may also be more appropriate for experienced learners who are improving existing skills rather than novices building new skills. Experienced learners already have mental models in place, so they don’t have the same needs for immediate correction as novices. They can get the benefit of delayed feedback.
Provide instructional feedback before retrying
I always use feedback before restarting a scenario. If a learner has reached an unsatisfactory ending in a scenario, it’s beneficial to do a short debrief of their decisions and what went wrong. Especially for more experienced learners, some of that feedback may be delayed from when they made the decision. You can summarize the feedback for several previous decisions on the path that led to the final decision.
This feedback should happen before they are faced with the same scenario decisions again. Otherwise, they could make the same mistakes again (reinforcing those mistakes) or simply guess without gaining understanding.
Thalheimer’s research also supports this.
“When learners get an answer wrong or practice a skill inappropriately, we ought to give them feedback before they attempt to re-answer the question or re-attempt the skill. This doesn’t necessarily mean that we should give them immediate feedback, but it does mean that we don’t want to delay feedback until after they are faced with additional retrieval opportunities.”
Will Thalheimer
Use immediate consequences with delayed feedback
In branching scenarios, we can use a combination of immediate intrinsic consequences (e.g., an angry customer response) and delayed instructional feedback (e.g., you didn’t acknowledge the customer’s feelings). Even if you don’t include immediate feedback or delayed consequences, the combination of these two elements is an effective approach.
Feedback before a retry or restart could count as delayed if it includes feedback for multiple decisions. If you let learners make 2 or 3 wrong choices before a restart, the combined feedback will effectively be delayed.
Mixed results from the research
I close with two more quotes from the research, this time from a 2023 article about medical education.
Indeed, there is evidence indicating that feedback delivered immediately after an item may optimise error correction (at least in some settings), whereas delayed feedback may enhance retention and transfer for correctly answered questions by providing an appropriately spaced second learning opportunity. Again, findings are mixed, and others have found delayed feedback to augment learning equally for both correctly and incorrectly answered questions. Interestingly, there may also be a greater advantage to delayed feedback when an initial question is more difficult.
That all sounds like it lines up with my recommendations above, right? But the article is titled “Timing’s not everything: Immediate and delayed feedback are equally beneficial for performance in formative multiple-choice testing.” That’s because they didn’t actually see the expected benefits of delayed feedback in their study.
Contrary to our hypotheses, the feedback timing effect was non-significant—there was no discernible difference between feedback delivered immediately versus delayed feedback.
This specific study is about formative multiple choice assessments in medical education, not about branching scenarios in workplace training. We should always be cautious about applying research from one context too broadly across other contexts. However, the practical reality is that there isn’t a ton of high-quality research about branching scenarios in workplace training. Sometimes, we have to look at the research in other areas and use the evidence to inform our designs.
This is a situation where personally, I think the evidence is strong enough to support these overall recommendations. But you shouldn’t take what I say as an absolute rule with no exceptions. You may find some situations where deviating from these design recommendations gets you better results, or that the feedback timing makes no discernible difference.
Looking for More?
Read more about branching scenarios and providing feedback.
Originally published 9/13/2017. Updated 9/7/2020. Updated with new research and images 5/1/2025.
