Context Clues Are Unreliable Predictors of Word Meanings

THREE STUDIES were conducted to determine the extent to which context helps students infer the meanings of unknown words. In Experiment 1, students in Grades 10 and 11 were randomly assigned to either a context or a no-context condition. The no-context group read lowfrequency words in isolation. The context group read the same words embedded in passages taken from novels. Experiment 2 was a repeated-measures study in which 39 students in Grade 11 read sets of words in isolation and also in passages taken from four different content areas. Experiment 3 was a systematic replication of Experiment 1 in which subjects were required to write definitions for the low-frequency words instead of choosing the definitions in a multiple-choice format. In none of the three experiments was there any statistically significant effect due to the context: t(99) = .552, p > .10; F(1, 38) = .227, p > .10; and t(83) = -.29, p > .10, respectively. The conclusion is drawn that instructional strategies that prioritize context clues should be reexamined.