Training never starts from zero

Training never begins with an empty mind.

Every employee arrives with previous experience, remembered rules, workplace habits, vocabulary, assumptions, and informal advice from other people. That prior knowledge becomes the structure through which new information is interpreted.

When it is accurate, relevant, and brought to mind at the right moment, it can accelerate learning. When it is incomplete, outdated, disconnected, or wrong, it can make good training harder to understand.

The practical implication is simple:

Before deciding how much content to provide, determine what knowledge the learner is bringing to the course.

Key takeaways

• Prior knowledge affects what learners notice, understand, remember, and apply.
• More prior knowledge is not automatically better; its accuracy, relevance, organization, and accessibility matter.
• A learner may possess useful knowledge without recognizing that it applies to the current task.
• A wrong answer is only one signal and does not prove that a stable misconception exists.
• Good training maps prerequisites, diagnoses the starting point, teaches selectively, and checks application in a new situation.

Prior knowledge changes the meaning of new information

Two employees can read the same policy and build different interpretations.

A new employee may struggle because the policy assumes vocabulary and organizational context that were never explained. An experienced employee may understand the words immediately but interpret them through an older procedure. A transferred employee may recognize the task while carrying rules from a previous organization that do not apply here.

This is why prior knowledge is more than a collection of facts. It includes relationships, expectations, habits, and judgments about when a rule applies.

Carnegie Mellon University summarizes the principle directly: prior knowledge can help or hinder learning. Accurate and appropriately activated knowledge creates a foundation, while knowledge that is insufficient, inaccurate, inert, or activated in the wrong context can interfere with learning. Source: Carnegie Mellon Eberly Center, “Learning Principles”.

A meta-analysis of domain-specific prior knowledge also found a meaningful relationship between what learners already know and what they learn next. The size of that relationship varied across studies and contexts, which is an important reminder that “prior knowledge” should not be treated as one simple quantity. Source: Simonsmeier and colleagues, “Domain-Specific Prior Knowledge and Learning”.

Prior knowledge can help, hinder, or remain unused

For practical workplace design, classify the learner’s starting knowledge into five states.

Accurate and connected knowledge supports the new task. The learner understands the relevant concepts and recognizes when to use them. Training can move more quickly toward complex cases and independent performance.

Accurate but disconnected knowledge exists but is not being applied. An employee may understand confidentiality yet fail to recognize that the same rule applies to an informal message sent from a personal phone. The problem is not absence of knowledge; it is failure to retrieve and connect it in context.

Incomplete knowledge covers part of the task but leaves out important conditions, exceptions, or consequences. An employee may know how to submit an incident report without knowing which events require one.

Outdated knowledge was once correct but no longer matches the current source or workflow. This is particularly risky because the learner may act quickly and confidently.

Incorrect knowledge is organized around a faulty relationship or rule. For example, a learner may believe that an event does not need to be documented when no one was injured, even though the organization’s current policy defines reportable incidents more broadly.

These states require different responses. Repeating the full course treats them as though they were the same problem.

Activating prior knowledge is useful only when it is relevant

A common recommendation is to begin training by asking learners what they already know. That can be useful, but activation is not automatically beneficial.

A systematic review of prior-knowledge activation identified many techniques, including open-ended prompts, visual representations, analogies, and activities before, during, or after reading. The review found that effectiveness varied partly with the amount, accuracy, and specificity of the knowledge being activated. Source: Hattan, Alexander, and Lupo, “Leveraging What Students Know”.

This matters in workplace training.

Asking experienced employees to recall “how we normally handle this” may activate an outdated shortcut. Asking learners for any personal experience related to a topic may surface an analogy that feels relevant but points toward the wrong rule.

Activation should therefore be guided.

Instead of asking, “What do you know about incident reporting?” ask a focused scenario question:

A client slips but says they are uninjured and refuses assistance. What should the employee do next, and which source supports that decision?

The question brings relevant knowledge forward while also revealing whether the learner is using the current rule, an older habit, or a guess.

Diagnose the starting point without turning it into a final exam

A useful diagnostic is short, low stakes, and connected to a design decision.

Use several forms of evidence:

• A scenario decision reveals which rule the learner applies.
• A short explanation reveals the reasoning behind the choice.
• A confidence rating shows how certain the learner feels.
• A task demonstration reveals whether procedural knowledge can be performed.
• A prerequisite check reveals whether the course is assuming missing vocabulary or concepts.
• A comparison across similar scenarios shows whether the reasoning is stable or context-dependent.

Do not infer too much from one response. A wrong answer may result from misreading, poor wording, a memory lapse, or a random selection. Even confidence does not automatically prove that a stable misconception exists.

The diagnostic should answer a practical question:

What should happen next for this learner?

Possible answers include skipping familiar material, reviewing a prerequisite, comparing an old and new procedure, receiving a worked example, or practicing a difficult decision.

Design from the learner’s state, not the document’s structure

Documents are normally organized for reference, governance, or completeness. Their section order is not necessarily the right learning sequence.

A better workflow is:

1. Define the workplace decision or performance.
2. Identify the knowledge and skills required.
3. Map the prerequisites.
4. Diagnose the learner’s starting state.
5. Provide only the explanation and practice needed.
6. Check a different but related application.

Consider a hypothetical incident-reporting policy.

The source may begin with definitions, then list responsibilities, reporting timelines, and form fields. A learning design should instead begin with the employee’s decision:

Does this event require immediate action, documentation, escalation, or more than one of these?

The prerequisite map might include the organization’s definition of an incident, the distinction between emergency response and documentation, role responsibilities, confidentiality, and reporting deadlines.

A beginner may need orientation, vocabulary, worked cases, and guided practice. An experienced employee may need only a change summary, edge cases, and a new scenario demonstrating that the updated rule has replaced the old one.

Represent prerequisites and conflicts explicitly

A knowledge graph can make prior-knowledge design more precise.

Useful relationships include:

• prerequisiteOf
• supportsUnderstandingOf
• conflictsWith
• supersedes
• appliesUnderCondition
• diagnosedBy
• practicedThrough
• evidencedBy
• derivedFrom

For example:

Current escalation procedure
supersedes
Previous escalation procedure

Scenario question 4
diagnoses
Ability to distinguish urgent response from routine reporting

Confidentiality principle
prerequisiteOf
Sharing incident information with an authorized person

This is relevant to LoreGraph’s work on turning source documents into structured workplace learning. The important design direction is not merely extracting sections from a document. It is preserving the relationships among sources, prerequisites, decisions, practice, and evidence.

Limits of the evidence

Much of the detailed prior-knowledge research comes from schools, universities, reading comprehension, science learning, and laboratory tasks. Workplace learners operate under different pressures: limited time, changing procedures, incentives, social norms, software constraints, and manager expectations.

The principles are useful, but they should be tested against actual employees and actual work.

Prior knowledge is also domain-specific. A person may be an expert in client communication and a beginner in the organization’s reporting system. Avoid assigning one permanent label such as “novice” or “expert” to the whole person.

Finally, a short diagnostic cannot reveal everything a learner knows. Treat it as evidence with uncertainty, not a complete psychological model.

Next step

Choose one existing course and write down its intended workplace outcome.

Then create three columns:

1. What learners must already know
2. What they may currently believe or do
3. What evidence would distinguish those starting states

Review the course again.

If it teaches everyone the same material without checking any of those conditions, add one low-stakes scenario before the first lesson and use the response to make one real instructional decision.

That small change moves training from content delivery toward learning design.

Sources and further reading

• Carnegie Mellon Eberly Center, “Learning Principles”.
• Hattan, Alexander, and Lupo, “Leveraging What Students Know to Make Sense of Texts”.
• Simonsmeier and colleagues, “Domain-Specific Prior Knowledge and Learning: A Meta-Analysis”.
• Shing and Brod, “Effects of Prior Knowledge on Memory”.
• National Academies, How People Learn II.