This paper offers a theoretical approach to vocabulary instruction from the evidence provided by lexical errors as the main sources of difficulty in the EFL acquisition process, it reviews previous research and from it suggests new ways of dealing with lexical errors in the classroom. Some practical implications are concluded which rely on lexical…

This note suggests delta method implementations for deriving confidence intervals for a latent mean effect size measure for the case of 2 independent populations. A hypothetical kindergarten reading example using these implementations is provided, as is supporting LISREL syntax. (Contains 1 table.)

Reports that good elementary school readers corrected more miscues than did poor readers, that poor readers relied more on graphophonemics to make corrections than did good readers, and that there were small differences between both groups' use of semantics and syntax to make corrections except as material increased in difficulty. (FL)