ABSTRACT
This study explores reader, word, and learning activity characteristics related to vocabulary learning for 202 fifth and sixth graders (N = 118 and 84, respectively) learning 16 words. Three measures of word knowledge were used: multiple-choice definition knowledge, self-report of meaning knowledge, and production of morphologically related words. Results indicate that significant vocabulary learning occurred on each measure and that certain words were easier to learn for certain types of readers. Controlling for other predictors in the model, reader characteristics like morphological awareness, reading comprehension, and language background were significant predictors of vocabulary learning, but not word reading fluency. Also, word characteristics like morphological family size and opaqueness were significant predictors of word difficulty but not number of morphemes or frequency of the word or root-word or affix. Controlling for other predictors in the model, morphological learning activities supported vocabulary learning for all 3 aspects of word knowledge. Implications for theory and instruction are discussed.
Acknowledgments
We thank the students and teachers who participated in the study.
Funding
This research was made possible through funding from Vanderbilt University’s Peabody College.
Notes
1 Scripts and Fidelity coding are available from the first author.
2 One intervention session was cancelled due to a fire drill. Students were taught in a different intervention session.
3 If odd, one of the lowest ranked three were randomly assigned to comparison instruction.
4 Source codes are available from the second author.
5 See Appendixes A and B for more information on model complexity results. Overall, results suggested that a one-parameter 3D model fit best with effects for word and teacher nestings included. Also, measurement invariance was confirmed, and standardized residuals for persons ranged from −0.135 to 0.117 and standardized residuals for items ranged from −0.036 to 0.072, suggesting good fit to the data.