Generative time telling in adults with disabilities: A matrix training approach

Abstract

Matrix training is a teaching strategy designed to facilitate generative responding through a process termed recombinative generalization. A skill that has yet to be addressed with matrix training is tacting time increments.Given the two-component features of time telling (i.e., hours and minutes) and the numerous teaching targets, matrix training was implemented for such behavior with two adults with developmental disabilities. This study used a 12 × 12 matrix that consisted of 144 time increments.Participants were taught 12 time increments, and probes were conducted to assess the remaining 132 untrained time increments. The participants responded correctly to approximately 50% of the untrained time increments during the post-assessment phase. Limitations and future research are discussed.

A systematic review of the matrix training literature

“…Alternatively, if learners are successful with recombining unknown components this could potentially increase instructional efficiency because it saves time teaching each unit to generativity. A few recent studies (i.e., Curiel et al., 2018, 2020) used a mixed presentation in which some stimuli on each of the axes were known and some were unknown—this approach could be considered for improving acquisition and generalization. To this end, future researchers might consider undertaking a line of basic or translational research as means to better understand the influence of variables suspected to affect recombinative generalization (e.g., whether stimuli are known or meaningful) and might be inspired by similar endeavors underway related to other types of generalization (e.g., stimulus generalization; Song et al., 2021).…”

“…For example, adding autoclitic frames or adding components, such as adjectives, to mands (e.g., combining the adjectives “big,” “small,” and “red” with the nouns “ball,” “animal,” and “car” to achieve multi‐component mands). A few researchers have shown success using matrix training to teach functional living skills (e.g., time; Curiel et al., 2020) and leisure skills (e.g., music; Langton et al., 2020). Expansions on these skills as well as more complex skills could be considered (e.g., money identification or foreign languages).…”

Comparing matrix sizes when teaching direction following to preschoolers with autism spectrum disorder

“…One intervention, matrix training, has been used to teach several socially significant and complex skills (e.g., direction following for pre‐literacy skills [Axe & Sainato, 2010], language‐building and communication skills [Curiel et al., 2016; Frampton et al., 2016; Goldstein & Mousetis, 1989; Jimenez‐Gomez et al., 2019; Karlan et al., 1982; Pauwels et al., 2015; Saunders et al., 2003; Wilson et al., 2016], sociodramatic play [Dauphin et al., 2004], telling time [Curiel et al., 2020], piano notes, and rhythms [Langton et al., 2020]). Matrix training increases the rate of acquisition by teaching the minimal number of combined targets specific to the matrix size while decreasing the time spent directly training all possible responses (Pauwels et al., 2015).…”

“…Prior studies have implemented matrix training with different size matrices. Most studies used 3 × 3 matrices (e.g., Dauphin et al., 2004; Frampton et al., 2016; Goldstein & Mousetis, 1989; Kohler & Malott, 2014; MacManus et al., 2015; Naoi et al., 2006), while other studies have implemented larger matrices (e.g., Axe & Sainato, 2010 [6 × 6]; Curiel et al., 2016 [5 × 4]; Curiel et al., 2020 [12 × 12]; Pauwels et al., 2015 [6 × 6]; Tanji & Noro, 2011 [7 × 7]; Wilson et al., 2016 [3 × 4]). To date, however, no identified study has compared the generalization outcomes produced by different‐sized matrices.…”

“…The existing literature on matrix training contains limited reporting on procedural integrity measures and maintenance data. Four studies to date have published procedural integrity data (Axe & Sainato, 2010; Curiel et al., 2016, 2020; Frampton et al., 2016). Moreover, only four studies reported collecting maintenance data one week to eight months after achieving target response mastery (Curiel et al., 2016, Dauphin et al., 2004; Kohler & Malott, 2014; Naoi et al., 2006).…”