Matrix training data should show the exact grid version, component status, taught and held-out cells, teaching trials, valid probes, independent responses, prompts, access failures, maintenance, transfer, and later direct teaching. Report raw counts beside percentages. Keep trained acquisition, untaught-combination performance, and system readiness in separate denominators so a summary cannot turn practice, excluded cells, or unavailable communication into apparent generalization.
Version the data record
Give each grid an effective date and list its rows, columns, excluded cells, response definition, teaching procedure, mastery rule, probe procedure, and change history. A new component, cue, or access method can create a new version.
Use separate measures
Taught-target acquisition uses taught trials or mastered cells. Untaught-combination performance uses held-out cells independently correct divided by held-out cells due. System readiness uses sessions with materials and access supports ready divided by sessions due.
Calculate yield transparently
Generalization yield can show untaught cells acquired divided by all cells acquired after the phase. The systematic review assessed acquisition, recombinative generalization, and maintenance separately. Report the raw number taught, emerged, and still unknown.
Anchor measurement decisions
The BACB outline covers measurement, experimental design, generalization, and interpretation. The tutorial illustrates matrix construction and trained versus untrained targets. The Ethics Code addresses documentation and evaluation.
A practical example
Maya's grid has 12 usable cells. Four are taught and mastered; six of eight held-out cells are independently correct. The report shows taught acquisition 4/4, untaught performance 6/8, and acquired-cell yield 6/10, with two cells still unknown.
Questions families can use
Which grid version produced the number? Are directly taught and untaught cells separate? Do exclusions have reasons? Is the denominator complete? Are maintenance, new materials, and partner changes separate tests?
Sources
- Behavior Analyst Certification Board, BCBA Test Content Outline, 6th edition
- Curiel and colleagues, A Tutorial for Implementing Matrix Training in Practice
- Curiel and colleagues, Matrix Training for Expanding the Communication of Toddlers and Preschoolers With Autism Spectrum Disorder
- Niland and colleagues, Comparing Matrix-Training Procedures with Children with Autism Spectrum Disorder
- Frampton and colleagues, The Use of Matrix Training to Promote Generative Language with Children with Autism
- Simeone and colleagues, Miniature Linguistic Systems for Individuals With Autism Spectrum Disorder: A Systematic Review and Meta-Analysis
- Behavior Analyst Certification Board, Ethics Code for Behavior Analysts
- American Speech-Language-Hearing Association, Augmentative and Alternative Communication
Finni resources
