Using encoding instruction to improve the reading and spelling performances of elementary students at risk for literacy difficulties

Weiser, Beverly, & Mathes, Patricia. (2011). Using encoding instruction to improve the reading and spelling performances of elementary students at risk for literacy difficulties: A best-evidence synthesis. Review of Educational Research, 81(2), 170-200.

This best-evidence synthesis encompasses 11 stringently filtered research studies that present evidence in favor of the explicit teaching of encoding skills (aka, “mapping” letters to the sounds they represent, especially in regularly spelled words) in early reading programs, especially those that serve high-need populations. Instruction in encoding is here shown to reciprocally (or even synergistically) work with decoding instruction. That is, teaching kids to put words together by associating letters with sounds (aka spelling) helps kids learn to decode words better when they are trying to read them, and vice versa. According to Weiser and Mathes, early reading programs do include decoding, but not always so much encoding, and they maintain that both skills should be emphasized for optimal benefit. The encoding interventions they focused on, and to which they limited their study inclusion, involved the manipulation and/or writing of words, and the target age level was mainly kindergarten through grade three, though studies with older readers were allowed if the students were reading below the third grade level.

According to Weiser and Mathes, best-evidence syntheses combine the best features of meta-analyses and traditional literature reviews. If theirs is a typical example, I agree. I found their no-nonsense, clear approach refreshing. The only dissonance I have going on here is the same one I usually have with meta-analyses. The goal in research syntheses seems to be to exclude studies rather than to include them. I realize that at times we want to employ the finest sieve possible to bring out the true “gold” in a body of research, rather than throwing out the widest nest possible. Still, there seems in these kinds of articles almost a glee in presenting criteria that can boil the list of studies down to nearly nothing. That always bothers me, and especially in this case, where three of the 11 included studies have the same author (Blachman) and one included study was done by the second author of this synthesis (Mathes). There may even be more “family relationships” than I am aware of within the included studies. Yes, we do need more studies (justifying those future studies is often the main point of writing a research synthesis), and I definitely think more are needed before we begin reshaping literacy curricula!

The research here is seen as support for the “connectivity” theories, which see language processes as interconnected, with development of multiple processes occurring in concert. Connectivity theories are something I’ve long believed got at some of the truth of what happens in literacy development, so that piece resonated for me. I guess for me the issue is not WHETHER to teach word encoding skills, but rather HOW we should teach such skills with young children. Explicit instruction is definitely needed sometimes, and it does not necessarily have to come packaged as “skill, drill, and kill.” My main fear, though, is that the findings of an article like this could be misused to justify (and, of course, to sell) meaningless skill instruction that will be detrimental to children’s motivation to learn.