This study marks the beginning of an attempt to discover something of the nature of the effect of reading on spelling. It was suggested by a previous investigation1 which showed that improvement in spelling takes place during the high-school and college years. Since systematic instruction in spelling is no longer given at these levels, what is to account for the improvement ? No doubt there are several operative factors. It would appear that a most important one is the learning which occurs as a by-product of reading. The present report covers an experiment with three hundred eighty men and women students of junior, senior and graduate university levels. It is concerned chiefly with three questions: Does reading for the purpose of being able to answer questions on the context result in any improvement in spelling for words encountered in the selection? Do good spellers surpass poor spellers in the amount of learning in spell ing acquired through reading? Is improvement in spelling effected by virtue of a slow reading rate? In preparation for the experiment a list of thirty-nine difficult words was selected from a number of books in which college students might be expected to be interested. Although no reference was made to word counts, no word was included which was considered entirely foreign to the vocabulary needs of university upperclassmen. To this list there were added eleven of the most difficult words from the Sixteen Spelling Scales.2 The fifty-word tests was administered to a trial group of one hundred forty-eight students. On the basis of their spellings the twenty most difficult words were identified. They were matched approximately for difficulty into two sets of ten words each. The general plan was to pre-test the students on both sets, present a reading selection in which were embedded the words of only one set, then post-test on both sets to compare the learning that had taken place. As a check against the possible influence of the inequality of difficulty
No takes yet. Share an insight, caveat, or question.
Luther C. Gilbert (1935) studied this question.