Peer learning is a very attractive option in getting both the most advanced and weakest learners extra learning. First, there simply are not enough adults in a classroom (at least not in my school) to give small group attention without underserving at least some other children. Two, this actually solves two issues in one. It keeps the most and least advanced children learning together at the same time, freeing the adult(s) to work intensely with the remaining children. Finally, there is a lovely cooperative aspect to such a design. I like the classroom culture and the ethos it represents.
The one thing that concerns me is whether it would work in 1st grade. I have noticed the behavior our book calls “horizontal decalage” and specifically the tendency that “any given child may have a whole variety of strategies and may use some or all of them on different problems” (Bee & Boyd, 2007, p. 166-167). In particular, on any given day or problem even the most accomplished students may employ an incorrect strategy out of his or her “wide variety of strategies.” We do math contests with groups at the end of the day for fun and learning. It is always fascinating to watch the children work in a group setting. Fascinating, but it is not terribly reassuring about how information transfers from the one with the most skill to the one with the least skill. Having said that, I would very much like to try this strategy and would love if it succeeds.
Bee, H., & Boyd, B. (2007). The developing child (11th ed.). Boston, MA: Pearson Education, Inc.