This first research question compared effect of three learning tasks, RC Task, GF Task, and PW Task, on incidental vocabulary acquisition based on immediate posttest.
The participants in each learning task scored a mean from 57.11 to 66.17 out of 90, the
49
full score, which indicates all these three contextualized learning tasks are helpful in the incidental vocabulary acquisition. This finding conforms to the statement from Huckin
& Coady (1999) that “text-based tasks can enhance incidental vocabulary acquisition.”
Nevertheless, the similar mean scores existing between RC Group and GF Group do not support the second assumption from the Involvement Load Hypothesis, “words
processed with higher involvement load will be retained better than words processed with lower involvement load” (Laufer & Hulstijn, 2001). On the other hand, the comparatively higher mean score from PW Group indicates PW Group with higher involvement load did bring more vocabulary learning than the tasks with lower involvement load. However, the mean score differences among the three tasks did not reach a significant level, which reveals involvement load differences existing among the three tasks may not be the sole factor that affects acquisition. The task with higher load in evaluation does not present its superiority over the one with lower load in evaluation.
The lackluster evidence for the Involvement Load Hypothesis from this study could be interpreted from three dimensions, frequency of occurrence, time on task, and language proficiency.
First, the similar vocabulary acquisition existing between GR Group and RC Group did not support the statement that vocabulary tasks can bring out more learning than reading only (Luppescue & Day,1993, and Knight,1994). The statistical evidence was also inconsistent with the prediction by the Involvement Load Hypothesis that the learning task with higher load, in this case two points for evaluation in GF Task should elicit more incidental vocabulary learning than one point for evaluation in RC Task, in account of the depth of processing. (Laufer & Hulstijn, 2001). This unexpected result might be attributed to the additional Chinese summary writing following the reading activity. This second activity may expose the participants to the target words in
productive way in addition to passively searching for the answers to the comprehension
50
questions. Thus, RC Task in this study might not be regarded as a reading only activity, but a reading plus task, which increases the frequency of vocabulary retrieval via a second access to the targets words in the text in a productive way for ESL learners (Min, 2008 & Paribakht & Wesche, 1997). This reading plus task may boost the effect and minimize the one plus difference in evaluation existing between RC Task and GF Task, so the participants in RC Task perform better than they are expected to be.
In GF Task, the participants also carried out two productive activities, filling in the blanks and summarizing. This one plus activity did not result in a better
performance because the matching task without feedback on accuracy impairs word learning. When participants misplace a target word, the word is not surrounded in the suitable context, and the word learning is interrupted. Therefore, the superiority of GF Task, with a higher load, was diminished by possible mismatching of meaning, thus eliminating the effect of one more plus of evaluation from GF Task than CR Task.
Second, the time constraint of the three learning tasks in this study was another reason needed to be taken into account. In this study, all the three learning tasks were conducted under a 10-min limit. The equal time period eliminated the influence coming from tasks, leading to the failure of PW Task with highest involvement load, 3 pluses of evaluation, in exceeding the two tasks, GF Task with 2 pluses and RC Task with one plus of evaluation. This time-controlled design did not meet Laufer’s (2001) suggestion that time is the essence of a task, needed not to be controlled when realizing a task.
When Laufer and Hulstijn (2001) proposed the Involvement Load Hypothesis, they explained word learning from the psychological aspects, need, search, and evaluation, but neglected the pedagogic aspect, the immersion time. When learners were immerged in a learning environment, the influence would increase with the passing of the time.
The longer the participants engaged themselves in the task, the more vocabulary could be acquired.
51
The finding that higher-involvement PW Task did not make a difference in vocabulary learning from the lower-involvement tasks did not necessarily contradict previous findings (Jing & Jianbin, 2009, Keating, 2008, Tu, 2004, and Hulstijn &
Laufer, 2001). Composition or sentences writing was more beneficial and conducive to language learners because the participants in the writing task in their studies consumed more time, named twice or three times as much time, than the task with lower load.
Therefore, when the time on task is held constant, the superiority of the task with higher load vanishes and is compensated by frequency of occurrence in other two tasks, which is also demonstrated by Flose (2006) and Hulstijn et al (1996). They found that, under the time constraint, a repeated task with lower load excelled a task with higher load.
That is to say, time on task was significantly associated with the vocabulary learning scores (Huang et al., 2012). When time on task is taken into account, the superiority of tasks with higher load will not hold (Keating, 2008).
Third, many previous studies (Hulstijn & Laufer, 2001;Laufer, 2003; Keating,2008;
Kim, 2008; Martinez-Frenandez, 2008; Yaqubi et al., 2010; Marmol &
Sanchex-Lafuente, 2013) on the Involvement Load Hypothesis were conducted among advanced or intermediate language adult learners. The adult learners with
high-proficiency may differ from teenager low-proficiency learners in three dimensions.
First, they have more metacognitive strategies to control over their learning process.
Secondly, they have more background knowledge to elaborate in text processing, be it recognition or production. Thirdly, they have broader and deeper vocabulary knowledge to begin with. With these knowledge sources, they can exercise more elaborative
learning when meeting a new word. Hence, depth of processing may not be determined only by the text that surrounds the vocabulary itself, but also in how much vocabulary, background, and metacognitive knowledge learners could bring in making meaning of words. Presumably, the more learner reservoir is summoned, as with adult
52
high-proficiency learners, the more profound the impact of task might be. The junior high students thus could not reach a depth of processing at all levels, as did the adult high-proficiency learners in above-mentioned studies. Their shallower level of
processing may be the cause of the minimal effect of task on vocabulary learning in this study, which is echoed by. Marmol and Sanchex-Lafuente’s study (2013) whose
experiment is on elementary school students.
In sum, the loading of a learning task weighed by the three components, need, search, and evaluation, did not help to anticipate the effect of a task. It is surmised that frequency of occurrence, time-spent-on-task, and learners’ knowledge may be also important elements on deciding the effectiveness and the efficacy of a task.
Interaction of Task and Proficiency Level on Incidental Vocabulary Acquisition
In immediate posttest, among different learning tasks, both the higher and lower levels acquired more scores from PW Task than RC Task, and followed by GF Task, but the differences among the three tasks are not large enough to be significant. However, if we look into the two levels respectively, we could find out some tiny differences
existing between them.
The higher-level proficiency learners benefited from the three learning tasks with less disparity than the lower-proficiency ones, although no interaction between Task and Proficiency Level was found. These similar attainments across tasks for
higher-proficiency learners indicated that learning tasks with different involvement loads facilitate the vocabulary learning equally for the higher-proficiency learners. In other words, the benefit by higher-proficiency learners from three learning tasks did not vary according to their involvement loads.
53
In comparison with higher-proficiency learners, participants with lower reading proficiency benefited more from PW Task than the lower-proficiency participants in GF Task and RC Task. The pictures provided in PW Task may be a factor promoting the performance of the PW lower-proficiency learners. A writing task is a laborious challenge or a mission impossible for the beginning learners, but pictures could reduce the processing load and make the task engaging for them. Unlike RC Task and GF Task, which command extensive text processing, PW Task does not require students to read an extensive text but only to write 6 sentences with the cue in the form of pictures for meaning. As stated in previous sections, text processing demands knowledge in three levels; pictures may reduce the need of such knowledge. In addition, the sentence writing task is less demanding in that it is a one sentence production. These factors may make up for the disadvantages of lower-proficiency level.
However, the lower-proficiency learners in GF Task and RC task performed almost the same. In other words, the loss of words is equal between the two groups, a mirror of higher proficiency learners. Again, the text in the two tasks may be difficult for the lower-proficiency learners to comprehend, leading to the inability to consolidate meaning in RC Task, and fill in blanks with correct words in GF Task, which further hampers the acquisition and thus retention of words. Reading the whole passage without other assistance may be a big challenge for them even when the target words are listed in the margin. A loss or a mismatch of meaning in the middle may block the move forward as the design of a task had expected. Hence, the vocabulary acquisition is disrupted and the benefit of deeper processing from a task with higher load fades out.