Matrix Training: Considerations for Recombinative Generalization and Efficiency of Acquisition
Mentor 1
Tiffany Kodak
Location
Union Wisconsin Room
Start Date
28-4-2017 1:30 PM
End Date
28-4-2017 4:00 PM
Description
Matrix training is a teaching procedure in which some combinations of component skills are taught and novel combinations are probed for emergence of recombinative generalization (i.e., the acquisition of novel combinations of previously mastered components). Matrix training aims to maximize instructional efficiency by arranging targets in a way that recombinative generalization is more likely to occur and thus eliminate the need to teach all possible combinations. The current study used an adapted alternating treatments design with generalization probes to compare two matrix training procedures and a procedure that involved teaching components directly. We evaluated procedures on their efficacy (i.e., acquisition to mastery) and efficiency (i.e., instructional time to mastery). The Components and Combination condition, a novel approach to matrix training, involved teaching individual component skills and the combination within the same trial and probing for novel combinations of the mastered components. The Combination Only condition, an approach in the extant literature, involved teaching the combination skills and probing for component skills and novel combinations. The third procedure, the Components Only condition, consistent with many early intervention practices, involved training each component separately and probing for the combination skill. One participant, Austin, was a 6-year-old male with ASD who learned noun-verb combinations in Spanish. For Austin, Components and Combination was the most efficacious and efficient teaching procedure. The second participant, Lexie, is a typically developing 4-year-old female who is learning to read consonant-vowel-consonant sight words, and her evaluation is still in progress. This study extends the matrix training literature by including a novel approach, the Components and Combination condition. Further, we compared the two matrix training approaches to a procedure that trained component skills in isolation. Results from this study will inform future researchers and practitioners on the most efficient and efficacious strategies to promote recombinative generalization.
Matrix Training: Considerations for Recombinative Generalization and Efficiency of Acquisition
Union Wisconsin Room
Matrix training is a teaching procedure in which some combinations of component skills are taught and novel combinations are probed for emergence of recombinative generalization (i.e., the acquisition of novel combinations of previously mastered components). Matrix training aims to maximize instructional efficiency by arranging targets in a way that recombinative generalization is more likely to occur and thus eliminate the need to teach all possible combinations. The current study used an adapted alternating treatments design with generalization probes to compare two matrix training procedures and a procedure that involved teaching components directly. We evaluated procedures on their efficacy (i.e., acquisition to mastery) and efficiency (i.e., instructional time to mastery). The Components and Combination condition, a novel approach to matrix training, involved teaching individual component skills and the combination within the same trial and probing for novel combinations of the mastered components. The Combination Only condition, an approach in the extant literature, involved teaching the combination skills and probing for component skills and novel combinations. The third procedure, the Components Only condition, consistent with many early intervention practices, involved training each component separately and probing for the combination skill. One participant, Austin, was a 6-year-old male with ASD who learned noun-verb combinations in Spanish. For Austin, Components and Combination was the most efficacious and efficient teaching procedure. The second participant, Lexie, is a typically developing 4-year-old female who is learning to read consonant-vowel-consonant sight words, and her evaluation is still in progress. This study extends the matrix training literature by including a novel approach, the Components and Combination condition. Further, we compared the two matrix training approaches to a procedure that trained component skills in isolation. Results from this study will inform future researchers and practitioners on the most efficient and efficacious strategies to promote recombinative generalization.
