Dissertations, Theses, and Capstone Projects

Date of Degree


Document Type


Degree Name



Educational Psychology


Linnea C. Ehri

Subject Categories

Cognitive Psychology | Educational Psychology


decoding skills; exposure to print; self-teaching; spelling; strategy instruction; vocabulary


Drake and Ehri (1984) showed that children could utilize a spelling pronunciation strategy in order to remember spellings of words. One purpose of the current study was to determine whether college students could also benefit from a spelling pronunciation strategy in remembering spellings of 20 commonly misspelled words. The second aim of the study was to examine the contribution of decoding skill, exposure to print and vocabulary knowledge in explaining variance in general spelling ability of college students. Based on Share's (1995) self-teaching hypothesis, each of these predictors was expected to explain unique variance in the ability to remember the spellings of words.

College students (N= 42) who were native speakers of English were recruited from an urban college. The mean age of participants was 22.5 (SD =7.87). There were 31 females and 11 males. The majority, 13 of them, were freshman who had not decided on their majors. An experimental design with pretest and posttest was adopted in order to measure the effects of a spelling pronunciation strategy. Half of the participants were trained to learn spellings of words by applying a spelling pronunciation strategy whereas the other half practiced reading the words. Results of immediate and delayed posttests showed a significant main effect of treatment. Participants who were trained by a spelling pronunciation strategy produced significantly more correct words, letters, silent letters, letters that represent schwa vowels, and double letters than the participants who practiced reading words (p <.001). Although there were no significant differences between the groups on the number of correctly spelled words on pretest, on posttest the participants who were trained by a spelling pronunciation strategy on average spelled 5.3 more words correctly than the participants who practiced reading words.

Hierarchical regression analyses showed that decoding and exposure to print explained significant variance in spelling ability if entered into the regression before vocabulary knowledge. However, when vocabulary was entered first, exposure to print and decoding did not explain significant additional variance in the model. One reason is that vocabulary shared substantial variance with decoding and exposure to print. When hierarchical regression analyses were conducted with vocabulary knowledge as the predicted variable and decoding and exposure to print as the predictors, results showed that both decoding and exposure to print explained significant unique variance not explained by the other predictor. Together they explained 37% of the variance in vocabulary knowledge. In turn the three predictors explained 42% of the variance in spelling ability.

These findings carry implications for spelling instruction. Students of every age can benefit by being taught how to create spellings pronunciations of complex words in order to remember how to spell the words. Limitations and directions for future research are discussed.