Research has shown that children demonstrate improvement in spelling performance a) when they are provided sufficient time for mastery, b) when they are allowed to compare their errors with correct spellings, and c) when practice is directed by students themselves rather than supervised by teachers. While research-based techniques offer sufficient cues or practice in error correction to help children master spelling, commercial spelling programs generally do not. Unfortunately, many teachers, despite their familiarity with research-based techniques, continue to rely on commercially
In the belief that teachers would be more inclined to use research-proven techniques if they were shown to be practical and effective in regular classroom settings, Maribeth Gettinger, University of Wisconsin/Madison, conducted a study in which she compared the spelling performance of students who used a commercial program with the performance of students who used an experimental, self-directed, error correction procedure.
Sixty-five children in three third-grade classrooms in a predominantly (89 percent) white, middle-class suburban school were enlisted for this study. All classes had been using the Lippencott Basic Spelling curriculum in which 15 words grouped by visual similarity and orthographic rules were presented each week. As part of this regular spelling instruction, teachers pretested students on the words each Monday. Twenty minutes of independent study were allocated on Tuesday, Wednesday and Thursday for the children to define the words, write them in a sentence and complete workbook pages. A final test was given each Friday.
The first of the three classes continued to use the Lippencott text. The second class used the same materials, however, the 15-word list was divided into three sets of five words each, from which students were asked to select five words each day for independent study on Tuesday, Wednesday and Thursday. The third class, also using three sets of five words, was taught to use an experimental, error-correction and practice procedure. Working in pairs, students in the third class tested each other on one set of five words each day and then corrected their own tests. For each incorrect word, they looked at the correct spelling and wrote a copy of their misspelled word, circling or highlighting the incorrect part. Next to this they wrote the correct spelling and circled the corrected part, repeating the word to themselves as they did this.
These students then practiced the word by looking at the correct spelling, repeating it and turning over their paper, writing the word and circling again the part that had been incorrect. Finally, after checking the spelling the paired students retested each other. This sequence of testing, error imitation and correction, practice and retesting was continued until all the words in the set were spelled correctly.
Data was collected on students1 performance for three spelling units of seven weeks each. For the first unit, all three classes used the Lippencott curriculum, averaging of 69, 72 and 66 percent correct on Friday tests. During the second unit, the third classroom used the experimental procedure, achieving 87 percent accuracy, significantly higher than either the first or the second class, in which students averaged 69 and 65 percent correct respectively.
Results from the third unit indicated that the experimental class continued to use the error correction procedure successfully, averaging 82 percent correct, compared to 67 and 68 percent correct in the other classrooms.
Spelling accuracy was also evaluated on the basis of the number of words spelled correctly in a story dictated at the end of each unit. Each story included 30 of the 90 words presented in the unit. In addition, every three weeks classroom teachers gave each student an overall rating for spelling performance on all written work. Overall, students in the experimental classroom earned higher scores once they had been taught the error correction procedure. In addition, data on the number of trials students needed to accurately spell the five words in each set indicated that as students became adept at the procedure, the number of learning trials decreased. During the first week of the study students needed an average of 2.4 trials in order to spell all words correctly, but by the sixth week they needed an average of only 1.6 trials. Gettinger concludes that the procedure not only enabled students to spell words more accurately, but seemed to increase the speed with which they learned to spell new words.
The results of this study show that imitating errors, modeling correct spelling and highlighting the incorrect parts of words appear to provide necessary and immediate feedback that improves spelling accuracy, especially for poor spellers. This self-directed correction procedure yielded higher weekly spelling test performance among third graders than either traditional spelling instruction or modified standard instruction in which the number of words to be learned was broken down into smaller sets for daily study. Moreover, the significantly higher performance of the error-correction class was maintained for the another six weeks following its introduction, suggesting that students continued to use the procedure effectively without further instruction. Gettinger concludes that this error-correction procedure can be used successfully in regular classrooms to improve spelling with no additional supervisory time required of teachers.
“Effects of Error Correction on Third Graders’ Spelling, Journal of Educational Research, Volume 87, Number 1, pp. 39-45.
Published in ERN, January/February 1994, Volume 7, Number 1.