Concept Mapping

Concept Mapping is a learning activity in which participants construct a visual diagram — a concept map — showing how the concepts in a domain relate to each other. Each concept appears as a labeled node. Each relationship between two concepts appears as a labeled connecting line. The act of deciding which concepts connect to which others and naming each relationship surfaces structural understanding that verbal explanation does not — which is why the method builds the kind of knowledge that transfers to new situations rather than only remains memorized. The labeled line is the method's distinguishing move: an unlabeled arrow is a node cloud, not a concept map.

The approach was developed by Joseph Novak at Cornell University in the 1970s, building directly on David Ausubel's Educational Psychology: A Cognitive View (1968) and its central claim about what Ausubel called meaningful learning — the learner incorporates new concepts into their existing knowledge structure by recognizing how the new concepts relate to concepts they already hold. Novak and Bob Gowin's Learning How to Learn (1984) documented the concept-mapping technique as the tool that made Ausubel's theory practicable in classrooms. Joseph Novak and Alberto Cañas's 2008 technical report The Theory Underlying Concept Maps and How to Construct and Use Them provides the contemporary reference, and the CmapTools software Novak's team developed has extended the method into digital-first workshop contexts.

Concept Mapping sits in the problem-and-inquiry family as the structural-visualization counterpart to Case Analysis. Where Case Analysis works backward through a resolved situation against a framework, Concept Mapping works structurally through a domain's conceptual relationships — what concepts exist, what relates to what, and how. Both activities produce pattern recognition; Concept Mapping produces pattern recognition about the domain's structure rather than about specific situations inside it.

What It Is

Four structural components define Concept Mapping.

• A defined domain or topic for mapping. The map has a scope — a specific domain, framework, problem, or system being represented. Unbounded domains produce cluttered maps that resolve nothing. The scope is named before mapping begins; it is usually one to three related concepts or an explicit central question the map will address.
• Concepts as labeled nodes. Each concept in the domain appears as a single labeled node — a word, a short phrase, a named idea. Concepts are the map's vocabulary. Participants either receive a starter set (for domains where vocabulary is standardized) or generate the concepts themselves (for domains where the mapping is also a diagnostic exercise).
• Relationships as labeled connecting lines. Each line between two concepts carries a label — a verb, a short phrase — that names how the two concepts relate. This is the activity's load-bearing element. An unlabeled arrow signals a relationship exists but does not specify it; a labeled line forces the participant to articulate what the connection actually is. Attention leads to encoding is a labeled relationship. An arrow from "attention" to "encoding" with no label is a wish, not a claim.
• Iteration through critique and refinement. Maps are not finished after the first draft. Participants review their own maps, receive peer critique, and revise — often several times. The revision is structurally part of the method: first drafts surface obvious concepts and connections; revision pushes the learner to question which relationships are actually supported and which are assumed.

When to Use It

Concept Mapping fits workshops where the target outcome is structural understanding of a domain — not only knowing the concepts but knowing how they relate, and being able to distinguish central from peripheral concepts and load-bearing from incidental relationships. It is especially strong for conceptually dense domains where verbal explanation loses the structure that a visual representation preserves.

Good candidates:

• Conceptually dense domains where relationships between concepts matter as much as the concepts themselves — frameworks with multiple interacting components, systems with dependencies, domains where participants routinely confuse similar concepts.
• Learning outcomes that include transferring structural understanding to new situations — participants who have mapped a domain can recognize analogous structures in different domains.
• Cohorts comfortable with visual thinking; cohorts new to visual methods may need a brief warm-up map on a familiar topic before mapping the workshop's real content.
• Sessions of 60 to 90 minutes — the initial map, peer critique, and revision each need time.

Less suitable:

• Domains with only a handful of linear concepts — the map reduces to a short chain and adds overhead without surfacing new understanding.
• Workshops where the outcome is procedural skill rather than structural understanding — Problem Sets or Case Analysis fit better.
• Cohorts who actively resist visual methods, where the overhead of the unfamiliar medium will outweigh the method's benefit.
• Very short sessions where iteration cannot happen — a single-draft concept map misses half the method.

In-Person and Virtual Delivery

Concept Mapping runs in both in-person and virtual delivery. The in-person form uses physical tools — paper, sticky notes, whiteboards — that let participants arrange and rearrange nodes freely. The virtual form uses digital canvas tools — Miro, FigJam, Mural, Lucidchart, or Novak's dedicated CmapTools — that support the same structural moves with different physical affordances.

In person. Participants work on large paper sheets or whiteboards, often with sticky notes as movable nodes and marker-drawn connecting lines. The physicality of rearranging sticky notes is itself part of the thinking — participants try relationships, see them, redo them. In-person delivery has a strong advantage in peer critique: maps posted on walls can be circulated through like a Gallery Walk, which produces cohort-wide pattern recognition across multiple mapped versions of the same domain. The workshop leader can see every map at once and identify where the cohort is converging versus where maps are diverging meaningfully.

Virtual. Participants use a digital canvas tool — Miro, FigJam, or similar — with one frame per participant. Nodes are drawn as shapes; connections are drawn as labeled arrows. The shared-canvas feature lets the workshop leader circulate across all maps without moving rooms, and the persistent digital artifact travels cleanly out of the session. Virtual delivery's advantage is the durability and editability of the maps — participants can continue refining their map after the session without recreating it from scratch. The tradeoff is that drawing labeled relationships in a digital tool is slower than sticky notes and markers for participants new to the platform, so the first ten to fifteen minutes often slip to tool-mastery rather than mapping.

Choosing the modality. In-person is slightly stronger for first-time Concept Mapping or sessions where the physical manipulability of nodes is part of the thinking. Virtual is stronger for multi-session programs where the map needs to travel across sessions, or for cohorts already fluent in a digital-canvas tool. For mixed-modality cohorts, run virtually with a shared canvas — the in-person participants lose the physical manipulability but gain the shared artifact. For single-session workshops with in-person-capable cohorts, in-person concept mapping usually produces deeper first-draft work.

How to Run It

Before the Session

1. Define the domain scope (leader design work). Name what the map will represent — a specific framework, a defined system, a central question and its related concepts. Scope tight: one to three related concepts as the map's core, not an entire field of study.
2. Decide whether to provide starter concepts or let participants generate them. For standardized vocabularies where the concepts are well-known (Kolb's four stages, the five phases of a framework), provide the starter list to avoid vocabulary confusion slowing the mapping. For diagnostic domains where participants' concept sets vary meaningfully, let them generate — the generation itself surfaces what each participant is holding.
3. Optional — distribute background reading. For domains where participants need shared vocabulary before they can map, send a short reading or glossary one to three days before the session. Concept Mapping works on the concepts participants know; they need enough foundation to have concepts to map.

During the Session

1. Frame the domain and the mapping task (5–8 minutes). Name the scope. Show an example of a simple concept map in a different domain so participants see what a labeled relationship looks like. Emphasize that every line gets a label — this is the rule that distinguishes Concept Mapping from arrow-diagramming.
2. Individual concept generation (8–10 minutes, if concepts are not pre-provided). Participants list the concepts they think belong in the domain's map, without yet trying to connect them. Listing first prevents premature structure.
3. Initial map construction (15–20 minutes). Each participant arranges their concepts spatially, draws connecting lines, and labels every line with how the two concepts relate. The leader circulates, watching for unlabeled lines (which have to be labeled) and forced connections (which have to be justified or removed).
4. Pair or small-group review (15–20 minutes). Participants exchange maps with one or two others. Reviewers read the map and name three things: relationships that feel vague or forced, concepts that seem missing, and connections that surprised them. Reviewers do not redraw the map; they raise questions for the author to consider.
5. Iteration and refinement (10–15 minutes). Participants revise their own map based on peer input — sharpening vague relationship labels, adding missing concepts, reconsidering connections they could not defend.
6. Full-group synthesis (10–15 minutes). The workshop leader surfaces patterns across the cohort's maps — which connections showed up across most maps (the load-bearing relationships), which concepts some had and others did not (the varying elements), where maps diverged meaningfully (the genuinely contested structural questions). The synthesis produces the cohort's shared structural read of the domain.

After the Session

1. Apply the map to a current project or decision (within one to two weeks). Participants take their refined map and use it to diagnose a real workshop, session, client engagement, or framework they are actively working on. The map becomes a structural lens: which concepts are present, which are missing, where the structure is load-bearing versus incidental.
2. Bring the applied map and the diagnosis to the next session (in multi-session cohorts). The cohort sees how the map survived contact with real work, which concepts turned out to matter more than expected, and where the map needed further revision under real-world pressure.

An Example in Practice

A 90-minute workshop for a cohort of ten coaches on designing demonstration sequences that participants can replicate. Each coach arrives with one demonstration they regularly use in their own sessions — a specific teaching moment where they model a skill or process for participants to then attempt.

The workshop leader defines the mapping scope: the structural components of effective demonstrations and how they relate to participant replication. She provides a starter set of eight concepts drawn from the research — framing, worked example, pace, chunking, scaffolding, observable behavior, near-transfer, feedback loop — and invites participants to add any concepts specific to their own demonstrations that are not in the starter set. Several coaches add concepts like permission to fail, prompt clarity, and participant readiness.

Initial map construction runs twenty minutes. Each coach arranges their concepts spatially and draws labeled connections. Some coaches put framing at the top with lines labeled prepares attention for → worked example. Others put observable behavior at the center with lines labeled is what participants will replicate during → near-transfer. Many struggle to label the line between pace and the other concepts — they can feel that pace relates to everything but cannot name how specifically.

Pair review runs fifteen minutes. Coaches exchange maps in pairs. The reviewers notice patterns: most maps have a line between framing and something else, but the relationship label varies wildly ("sets up," "introduces," "prepares attention for," "provides context for"). The diversity of labels is itself a diagnostic — the cohort does not yet share a precise view of what framing actually does in a demonstration. Reviewers flag forced connections, vague labels, and missing concepts.

Iteration runs twelve minutes. Coaches sharpen their labels and add concepts they realized were missing.

Full-group synthesis runs fifteen minutes. The leader surfaces the patterns: almost every map has observable behavior as a central node (confirming its structural importance); the pace concept showed up on most maps but was poorly connected on most (suggesting the cohort knows pace matters but has not structured its relationship to the rest of the demonstration). The coaches leave with a sharper structural view of their demonstrations and a specific direction for refinement — not "make better demos," but tighten the relationship between pace and the observable behavior being modeled.

Variations

• Collaborative Concept Mapping. A small group of three or four builds one shared map together rather than each participant building their own. Used when the domain is collectively owned (a team working on a shared framework) or when the cohort would benefit from the explicit negotiation that collaborative mapping requires — every relationship label has to be agreed by the group.
• Linking Concept Map. Starts with a single central concept and builds outward, adding directly-related concepts first, then concepts related to those, and so on. Produces a hierarchical map with clear central-versus-peripheral structure. Used when the domain has a clear core concept and the mapping question is what is central versus what is peripheral.
• Comparative Concept Mapping. Two concept maps on related domains are constructed and compared structurally — what concepts appear in both, what relationships differ, where the structures overlap and where they diverge. Used when the learning goal is seeing structural parallels or differences between two domains (for example, mapping client onboarding in a short program versus a year-long program).
• Digital Concept Mapping. The map is constructed in a digital canvas tool (Miro, FigJam, Lucidchart, CmapTools). The structural moves are the same; the physical affordances differ. Used for virtual delivery, multi-session programs where the map needs to persist and evolve across sessions, or cohorts already fluent in digital-canvas tools.

Which Method It Serves

Concept Mapping is primarily an Inquiry-Based Learning activity. The mapping process is itself an inquiry into the domain's structural relationships — participants investigate what connects to what and why, and the investigation produces understanding they could not have arrived at through direct instruction. The method activates every element of inquiry-based methodology: an open structural question (how do these concepts relate?), hypothesis formation (each proposed connection is a claim about the domain), evidence examination (peer critique pressure-tests claims), and synthesis (the refined map).

Concept Mapping also functions inside other methods. In Problem-Based Learning, concept mapping can surface the structural elements of a problem before participants work on solving it — a map of the problem's concepts clarifies what is actually at stake. In Case-Based Learning, a concept map of the case's elements can be built as part of the analysis, which surfaces structural patterns analytical-framework-based reading can miss. In Project-Based Learning, a concept map of the artifact's structure can be built before the first draft — participants design the structure before they build it.

Design Considerations

• The quality of the relationship labeling determines the value of the activity. Concept Mapping is defined by the labeled connecting line. An unlabeled arrow between two concepts is a node cloud, not a concept map. Every line must carry a label — a verb, a short phrase — that names how the two concepts relate. Generic labels like "relates to" or "connects with" are signs the participant has not actually articulated the relationship; specific labels like "prepares attention for," "is a precondition of," "reduces load on," "triggers," or "constrains the scope of" force the participant to name the structural claim they are making. Before running Concept Mapping, pressure-test the instruction: is the requirement for labeled relationships named clearly enough that participants will not default to arrow-drawing, and does the leader's own example map demonstrate what meaningful labels look like? If either answer is soft, the activity's signature move — which is what separates Concept Mapping from general diagramming — will get lost.
• Scope the domain tightly. Unbounded domains produce cluttered maps that do not resolve any structural question. One to three related concepts as the map's core is the practical range. Larger domains should be decomposed — map the subdomains separately, then compare or compose the sub-maps if cross-domain structure is itself the target.
• Keep individual mapping as the first step before peer critique. Participants who see a peer's map before they draft their own tend to anchor on that peer's structure, which loses the diagnostic value of seeing how differently participants structure the same domain. The individual drafting stage is what makes the peer critique productive — reviewers are comparing their map to the author's rather than importing the author's structure wholesale.
• Design the peer critique around structural questions, not redrawing. The reviewer's job is to raise structural questions — what relationships are vague, what concepts seem missing, what connections look forced — not to redraw the map. Reviewers who edit the map are producing their own map on top of the author's; reviewers who ask structural questions are strengthening the author's thinking. Frame the critique explicitly: what three questions would you raise about this map?
• Build in revision time, not just drafting time. A single-draft concept map captures the participant's first-pass structural view, which is usually the shallow one. The revision stage — after peer critique — is where the map becomes genuinely thought through. Sessions that compress the revision stage produce drafts labeled as maps; sessions that protect the revision stage produce real concept maps.

Where this goes next: Concept Mapping constructs structural understanding of a domain through labeled relationships. The next activity in this family uses structured hypothesis-formation and testing to investigate an open empirical or analytical question. See Hypothesis Generation.

Related

Read this next:

• Hypothesis Generation — the next activity in the problem-and-inquiry family; moves from mapping structural relationships to generating and testing hypotheses about them.

Foundations of this activity:

• Inquiry-Based Learning — the primary method Concept Mapping serves; the mapping process is itself an inquiry into structural relationships.
• Case Analysis — the retrospective-analysis counterpart in this family; compare Case Analysis's situation-specific patterns with Concept Mapping's domain-wide structure.
• Scaffolding and the Zone of Proximal Development — the Foundations article that touches Ausubel's meaningful-learning theory, the cognitive basis Concept Mapping operationalizes.
• The Guiding Principles of Active Learning — the principles Concept Mapping activates, especially honest reflection and focused outcomes.

Where this leads:

• Which Active Learning Method Fits Your Workshop? — Concept Mapping is the strongest sub-activity when structural understanding of a domain is a target outcome.
• Project-Based Learning — the method where Concept Mapping most often appears as pre-build design work before artifact production.

References

Ausubel, D. P. (1968). Educational psychology: A cognitive view. Holt, Rinehart and Winston.

Novak, J. D., & Cañas, A. J. (2008). The theory underlying concept maps and how to construct and use them (Technical Report IHMC CmapTools 2006-01 Rev 01-2008). Florida Institute for Human and Machine Cognition.

Novak, J. D., & Gowin, D. B. (1984). Learning how to learn. Cambridge University Press.