Variable-by-Variable Matrix

Description

Cross the ordered indicators of two main variables and enter case codes in the cells, to see which kinds of cases combine which indicators.

Used when a few major variables underlie many specific phenomena and you want to see how they connect without losing the specific indicators. Each variable's indicators are defined and ordered, with colleagues adjudicating the order: in the source example problem types ran from most to least tractable as columns, and 23 coping strategies were sorted into nine clusters ordered from shallow to deep as rows. Cases are ordered on an outcome and given codes. Cells hold the codes of cases where that pairing was observed, with '?' when the link is uncertain, and a case may appear several times. Pairs are entered in a cross-case analysis meeting from displays, field notes and researcher judgement, with differences settled by discussion. In the example, shallow coping appeared mostly in less successful schools and deep coping mostly in successful ones. The matrix is only as good as the indicator orderings.

Attributes

Loss type
Siloed
Effort
Days

Steps

  1. Define the indicators of each variable and order them, with colleague feedback.
  2. Order the cases on an outcome and give them codes (for example by level of success).
  3. Make one variable's indicators the rows and the other's the columns.
  4. Enter case codes in the cells for each observed pairing, adding '?' where the link is uncertain; allow several entries per case.
  5. Read where each group of cases falls in the matrix.
  6. Check the patterns against case data before drawing conclusions.

Sources

  • Miles, M. & Huberman, A. M., Qualitative Data Analysis: An Expanded Sourcebook, 2nd ed. (1994)

Also known as

  • Indicator-by-indicator case matrix
  • Problem-by-coping matrix

Categories

  • User Research
  • Qualitative Analysis & Synthesis