Interaction Matrix Table¶
Artifact — instantiates Interaction Effect Mapping
A table or grid used to display factor combinations, combined effects, evidence confidence, and recommended actions.
An Interaction Matrix Table is the shared, human-readable artifact that holds the map — a factor-by-factor grid in which each cell records what a specific pairing does, how confident that knowledge is, and what to do about it. Its defining property is that it is a record and a display, not a test or an analysis: it does not generate evidence, it curates and arranges it so a team can reason about the whole combination space at once. Every populated cell carries four things — the classified interaction type (reinforce / cancel / interfere / neutral / conditional), the effect direction and size, an evidence-confidence mark, and a recommended design action. Reading across a row shows how one factor behaves against all its partners; scanning the confidence marks shows where the map is solid and where it is guesswork. The table's discipline is that it never lets a bare observation sit without a classification and an action attached.
Example¶
A specialty coffee roaster is building its next tasting menu and needs to know which pairings of variables actually improve the cup. Three factors matter: bean origin (Ethiopian / Colombian / Sumatran), roast level (light / medium / dark), and brew method (pour-over / espresso / cold brew). The head roaster lays out a matrix — origins down the side, a roast×brew combination across the top — and fills each cell from blind cupping sessions.
Some cells read reinforce: Ethiopian + light + pour-over scores far above what either variable predicts alone, marked high-confidence from repeated tastings, action feature on the menu. Others read interfere: Sumatran + light + espresso is thin and sour, high confidence, action drop. One cell — Colombian + dark + cold brew — is promising but rated low-confidence because only one panel tasted it; the recommended action is re-test before committing, not adopt. The finished table becomes the menu-design conversation: the team argues from the grid, not from whoever tasted most recently, and the confidence column keeps a single lucky cup from being promoted to house blend.
How it works¶
- Lay out the grid. Choose the factor axes and render every combination of interest as an addressable cell — the structured relation the archetype calls for.
- Populate each cell with four fields. Classified interaction type, effect direction/size, confidence mark, recommended action; a cell missing any field is visibly incomplete.
- Encode confidence and action visibly. Use color, symbols, or a confidence column so strength of evidence and the design response are scannable, not buried.
- Keep it a living record. As new evidence lands, cells are revised; the table is the durable memory of what the team knows about combinations.
Tuning parameters¶
- Cell granularity — one cell per pair, or nested cells per condition/context. Finer cells capture conditionality but crowd the grid.
- Confidence scheme — a three-tier tested/suspected/guessed mark versus a numeric interval. Richer schemes carry more nuance but demand more discipline to keep honest.
- Action vocabulary — the fixed set of recommended responses a cell may carry (combine / separate / sequence / monitor / re-test / drop); a tighter vocabulary makes the table decisive, a looser one flexible.
- Symmetry — whether the table treats A-with-B and B-with-A as one cell or two (relevant when order or direction matters).
- Empty-cell policy — whether untested combinations are shown blank, greyed, or omitted; showing them keeps the coverage gaps visible.
When it helps, and when it misleads¶
Its strength is that it makes the whole combination space legible in one place and forces every observation to arrive already classified, confidence-rated, and tied to an action — which is what turns a pile of test results into a decision surface a group can share.[n1] It is the artifact that keeps interaction knowledge from living only in one expert's head.
Its failure mode is that the table is only the representation: a neat, fully-colored grid can project false completeness, especially when low-evidence cells are colored as boldly as tested ones. The classic misuse is treating the artifact as the analysis — filling cells with plausible guesses and then acting on them as if they were measured, or freezing a grid that has quietly gone stale. The guarding discipline is to keep the confidence field load-bearing (a cell's color must be earned by its evidence, not its plausibility) and to mark untested cells as gaps rather than hiding them.
How it implements the components¶
interaction_matrix— is the structured factor-by-factor relation object, in its tabular display form.interaction_type_classification— each cell carries a classified interaction label, not a raw number.design_update_rule— every cell holds a recommended action, linking the map to a design response.evidence_confidence_rating— a confidence mark per cell distinguishes tested from suspected relations on the face of the artifact.
It does not implement factor_list, context_stratification, or review_or_iteration_rule — the node set, the load-regime strata, and the propagation-pathway view belong to its nearest twin, Dependency Interaction Map, which represents relations as a directed network of dependency paths rather than as a factor×factor grid.
Related¶
- Instantiates: Interaction Effect Mapping — is the durable artifact the mapping logic writes into.
- Consumes: Factorial Experiment — its measured, classified interaction terms populate the table's cells (as do the other testing and analysis mechanisms).
- Sibling mechanisms: Factorial Experiment · Pairwise Combination Testing · Compatibility Screening · Dependency Interaction Map · Treatment Interaction Analysis · Design of Experiments Protocol
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: Interaction Matrix Table operates as a durable record, ledger, register, or trace whose value depends on preserving actual state or history because it a table or grid used to display factor combinations, combined effects, evidence confidence, and recommended actions
Independent corroboration: The frozen evidence defines Interaction Matrix Table as 'A table or grid used to display factor combinations, combined effects, evidence confidence, and recommended actions', so its operative form is Record, Log & Register.
Nearest alternative: Representation, Specification & Plan — The living table is retained as the current durable memory of combination evidence and actions rather than a one-time specification.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Futurism & Strategic Foresight
Origin pattern: Convergent development
Present-day reach: Multi-domain
Rationale: Cross-impact matrices were formalized in forecasting and strategic-foresight practice to record how events or factors modify one another.
Related originating lineages:
- Statistics & Experimental Design — Factorial design and interaction-effect analysis materially shape evidence and confidence for combined factor cells.
Review resolution: Both independent reviews place the primary lineage in futurism_foresight. The queued differences (reported_ambiguity) concern secondary metadata rather than primary provenance. The final retains statistics_experimental_design only where a reviewer supplied a formative-lineage rationale; this does not convert downstream applicability into origin. origin_mode=convergent because the reviewers document independently established or materially co-developing traditions. domain_reach=multi_domain records application breadth separately from provenance.
Attribution caveat: The generic table can instantiate either qualitative cross-impact analysis or quantitative interaction analysis.
Review outcome: Reconciled after independent review; medium confidence.
Notes¶
[n1] Cross-impact analysis (Gordon and Helmer, 1960s) pioneered the use of an interaction matrix in forecasting — a grid recording how each factor or event modifies the others. The enduring lesson is that the matrix organizes and communicates interaction knowledge but does not itself establish it; the entries are only as good as the evidence behind them. ↩