Define the objective of using the Matrix Diagram. This may be a statement such as, 'Focus design improvements on key customer requirements', which will be used later to direct activities.
Recruit a team who have the time and knowledge to work on achieving the objective. Building a Matrix Diagram can take a lot more effort than some of the other diagrams described in this book, and a longer term commitment may be required.
Decide what needs to be compared to achieve the objective. This will result in two or more lists being identified where the investigation of their relationships will help to achieve the objective.
This might also include identification of criteria to help decide what should and should not be included in the list. For example, if comparing insects with diseases, one criterion may be to exclude any insects which are unlikely to appear in the geographical area of study.
Identify the appropriate matrix to use. In approximate order of common use, these are:
The L-matrix is by far the most common diagram. If there are more than two lists, then a set of L-matrices may still be the best approach, unless the additional relationship mapping given by other matrices is required.
The T-matrix is useful when there are two distinct sets of questions about a core list, for example comparing school subjects against students and against teachers. An indirect relationship can be inferred between the two side lists.
The Y-matrix closes the loop on the T-matrix, and is useful for comparing three tightly coupled lists. It can also be used as a practical simplification of the C-matrix.
The X-matrix is useful for comparing two pairs of complementary lists, with each pair occupying diagonally opposite lists (as they have nothing in common and need not be compared). For example comparing men and women against activities in athletic and intellectual pastimes, with men and women opposite.
The C-matrix compares three lists simultaneously, such as the people, products and processes in a factory. Being three-dimensional, it is difficult and complex to produce and draw. It becomes easier if there are few relationships to map.
Decide how list items are to be compared. The most common is the strong/medium/weak relationship, although there may well be circumstances where other relationships may be more appropriate. For example, when comparing a list of people against tasks, the comparison may indicate prime responsibility, influence and interest.
This stage may also include identification of symbols to use. The most common symbols are as shown in the diagram above, although any other symbol set may be chosen. Although symbols are easier to interpret visually, numbers may be preferable, particularly if a computer spreadsheet is being used.
Derive the lists, using guidelines from step 3. Individual items may be easily available, or may require significant effort to acquire, for example when determining key customer requirements.
Beware of long lists resulting in large and unwieldy matrices. Initial lists may need to be trimmed down for an early focus on key areas.
Perform the comparison of the matrices, consistently using the rules defined in step 5. It is often preferable to aim for a relatively sparse matrix than to identify even the weakest relationship, as a symbol in every cell can result in key relationships being difficult to spot.
Evaluate the final matrix, looking for items of significance which will result in specific actions being carried forwards. Things to look for include:
Unimportant items which have few or no relationships with the other lists.
Key items which relate to many of the items in the other lists.
Patterns which strike you as odd, and which may bear further investigation.