Read on Twitter
Happy about the #RRQ paper but don't want to make it look like we are just simply advocating WM training. In the paper, we discussed more nuanced and different working memory training frameworks through the lens of cognitive element model of reading instruction 
https://twitter.com/PengPeng_UT/status/1276194765743235072
https://twitter.com/PengPeng_UT/status/1276194765743235072
Pretty clear by now domain-gen WM training on education irrelevant tasks (e.g., click rocks on a screen) has no far transfer on academic outcome. But, evidence is ROBUST across metas on near transfer to outcomes sharing similar structure/strategy use with the training tasks
Having improvement on WM tasks (often very challenging) is a BIG deal and accomplishment, especially for at-risk children. So, such near transfer effects should NOT be ignored and should be considered useful in thinking about helping at-risk readers who often have WM deficits
One thought we have is, why not make the reading outcome a "near transfer" task that is similar to the training tasks with WM loading. Indeed, when looking at reading activities from sound blending early on to reading comprehension, working memory is all over the place.
One such near transfer and WM training example is retell (see the detailed description in the pic).
We call it domain-specific WM training, in line with LT-WM model (Ericsson, 1995) and WM-Reading development model (Peng et al., 2018). To increase near transfer, WM training should take place in reading processes with explicit instruction, as emphasized by Science of Reading
An increasing number of rigorous interventions on typical and at-risk readers adopted such domain-specific WM training paradigm and showed moder-to-large effects on reading, even against skill-based controls (Carretti 2014, García 2013, Fuchs 2017, Peng 2017, Goodrich 2020).
So, we think the story of WM training is not over yet, at least for special educators who want to find the best practice for non-responders with cognitive deficits. More research is needed for the domain-specific WM training to figure out the nuanced task design and instruction
Peng, P., & Goodrich, M. (2020). The cognitive element model of reading instruction. Reading Research Quarterly. https://www.researchgate.net/publication/342281780_The_cognitive_element_model_of_reading_instruction
