Search

The paper reports on a study among primary-school students of the process of deriving word meaning from written context in a first language. A sequential analysis of think-aloud protocols revealed that the students inferred one or more meanings, checked their inferences and then rejected or accepted them. These activities were performed in a highly flexible manner and their order varied. Four major sequences, which were not equally effective, showed up in a cluster analysis. The paper concludes by discussing implications for instruction and think-aloud research.

A meta-analysis of 21 instructional treatments for improving the skill of intentionally deriving word meaning from context during reading indicated a medium effect size of 0.43 standard deviation units. An exploratory multilevel regression analysis revealed that clue instruction seems to be more effective than other instruction types or just practice. Effect size correlated negatively with class size.