REAP Study on Focusing of Attention (Spring 2007)
REAP Study on Focusing of Attention
This study examined the effect of highlighting target vocabulary words in practice readings for vocabulary practice. Previous research has examined the highlighting of words to facilitate incidental acquisition (incidental because the main task was reading comprehension). This work, however, was aimed at studying the effects of highlighting on intentional vocabulary learning. The purpose of students using the REAP tutor is to learn vocabulary words from reading materials and subsequent practice exercises.
When reading a text in REAP, students know that they should pay attention to unknown vocabulary. When words are highlighted (as they have been in previous iterations of the REAP tutor), students can simply access the definitions for highlighted words and not actually read any of the context. However, if the words are not highlighted, students must read more of the text to find the target vocabulary words. Therefore, not highlighting words might encourage students to more deeply process materials.
On the other hand, not highlighting words may lead to students not noticing the words. Conscious noticing () of unknown words has been proposed as an important component of vocabulary acquisition.
This study was a pilot study for a study to be performed in Spring 2007 in the same class. This was the first time REAP was used by ELI students in a class other than Reading 4. Students only used the system for approximately one month and the number of subjects was fairly small.
Does highlighting target words improve learning in a reading task aimed at the intentional vocabulary acquisition of vocabulary?
Post-test performance on cloze questions for words identified as unknown on a pre-test and practiced using REAP.
Number of words practiced.
Overall post-test scores.
Sentence production, long term retention, use of practiced words in writing for other classes.
Highlighting of target vocabulary words in practice readings.
Not highlighting target words will cause students to spend longer on each text. The deeper processing of texts will not, however, outweigh the smaller number of practice opportunities available resulting from seeing fewer texts over the course of the study. Time on task is the same, so students spending longer will see fewer texts. Overall post-test scores will thus be higher for the control condition in which target words are highlighted.
Students were randomly assigned to control and treatment groups. In the control group, as defined here, the target vocabulary words were not highlighted (or linked to definitions) in the readings. In the treatment group, target words were highlighted and linked to dictionary definitions of the words.
There were 27 students who participated in the study. Only 18 came to class on the day of the post-test. Unfortunately, due to absences, there was a significant difference in the time on task between the control and experimental groups (Control Group: N=10, M=163.95 minutes SD=98.7 min; Treatment Group:N=8, M=82.9 min, SD=46.2 min; p=0.49).
The number of target words looked up in an online dictionary (either by clicking or by typing them into an HTML form element) was statistically significantly different between groups (p<0.001). Students for whom the target words were highlighted looked up 2.26 words per reading, while those students without highlighting only looked up 0.42 target words per reading. In contrast, the mean number of non-target words looked up (none of which were highlighted) was essentially equal (3.75 for control, 3.69 for treatment).
Students with highlighting of target words also performed significantly (p=0.001) better on post-reading vocabulary exercises (M=95% correct) than did students without highlighting (M=56%). However, students in both groups performed approximately equally well on the post-test (M=52.9% correct for control, 59.1% for treatment with highlighting), and the difference between groups was non-significant (p=0.22).
It seems therefore, that the short term performance gains for students in the highlighting condition that were seen on post-reading practice exercises did not lead to robust learning of vocabulary, as measured on the post-test. However, further study is warranted because of the small sample of students and differences for time on task. One note on time on task: the post-test measurements given here are only for words seen in readings. While students who spend more time with the tutor likely saw more target words, it is less likely that they saw each individual word much more often than did the students in the control group.
All t-tests for significance were two-tailed and for independent samples.
Analysis of results pending.
De Ridder, I. (2002). Visible or Invisible Links: Does the Highlighting of Hyperlinks Affect Incidental Vocabulary Learning, Text Comprehension, and the Reading Process? Language, Learning & Technology, Vol. 6, 2002