In 2 experiments, independent measures of detection and correction of
text errors were examined. In Session 1, Ss tried to detect spelling,
grammatical, and meaning errors in a text; in Session 2, Ss tried to c
orrect the same errors in workbooklike exercises. Errors were classifi
ed according to whether they were detected in Session 1 and whether th
ey were corrected in Session 2. With the exception of spelling errors,
knowledge of how to correct an error was necessary but not sufficient
for detection. For spelling error detection, knowledge of how to corr
ect may be both necessary and sufficient. In Experiment 2, detection o
f meaning errors was manipulated by having Ss answer comprehension que
stions, surface questions, or no questions. Both question groups detec
ted nearly twice as many meaning errors as the no-question group. Ques
tioning had no effects on detection of spelling or grammatical errors.