Participants in the study were students from four classes in Changhua Senior

Chapter Three Methodology

This chapter is about the method in the present study. It consists of four parts,

including the participants, instruments, procedures, and data collection and analysis.

3.1 Participants

Participants in the study were students from four classes in Changhua Senior

High School, including two classes in the first grade and two classes in the third grade.

The students from the first grade were categorized as participants of low proficiency

and the students from the third grade were classified as participants of high proficiency

in the study. All of the participants were male students, aged from seventeen to

nineteen. As far as the participants’academic performance was concerned, Changhua

Senior High School ranked at the top among all senior high schools in central Taiwan.

Thus, their English proficiency was above the average among students of their age in

Taiwan. Regarding time invested in English, students in the first grade attended five

English classes each week, while students in the third grade had seven English classes

each week.

3.2 Instruments

The instruments used in the study included a pre-test of target words, two

vocabulary immediate post-tests and a delayed post-test.

3.2.1 A pilot study for the target words

To ensure that the participants in the study have little knowledge on the target

words, a pilot vocabulary test was conducted. First, a class of the third grade,

consisting of forty five students with better academic performance than the

participants in the study, was chosen. They were then required to write down L1

translation of thirty-two words adopted from the two experimental reading passages.

As Table 3-1 illustrated, the results of the test confirmed that learners had little

knowledge on the twenty target words chosen in the present study.

3.2.2 Pre-test, Immediate Post-test and Delayed Post-test of Target words

The vocabulary pre-test (see Appendix A) consisted of twenty multiple-choice

questions in sentence context. Each of the two vocabulary immediate post-tests

contained ten multiple-choice questions picked out from the pre-test and arranged in

different order. The delayed post-test contained the same questions as those in the

pre-test but presented in different order. The different order of the questions on the

target words in the pre-test, the immediate post-tests and the delayed post-test was to

exclude any possible effects of episodic memory for the order of presentation.

3.2.3 Reliability and Validity of the Vocabulary Tests

To ensure reliability of the vocabulary test, I conducted a pilot test by having one

class, who were not participants of the study, take the same vocabulary test twice with

a two-week interval. After comparing the results of the test, similar performance was

found and the reliability of the vocabulary test was attested.

To ensure validity of the vocabulary test, I had the test checked by one

experienced teacher, who agreed that the design of the test actually measured the

vocabulary knowledge on the target words in the reading tasks.

Table 3- 1 The results of the pilot vocabulary test Vocabulary Number of correct

translations (of forty five students)

Target words Non-Target words

Contemporary 4 √

Ethical 20 √

Criterion 1 √

Emphatic 5 √

Presume 3 √

Incentive 0 √

Overcome 38 √

Faculty 0 √

Relish 0 √

Ponder 5 √

Integrate 1 √

Generalize 3 √

Stimuli 4 √

Priority 22 √

Deficiency 0 √

Render 0 √

Dual 2 √

Core 8 √

Preside 0 √

Filter 0 √

Preliminary 0 √

Proposal 9 √

Smoothly 36 √

Diagram 3 √

Resolute 3 √

Friction 3 √

Dissolve 4 √

Condense 2 √

Radius 0 √

Density 10 √

Patch 0 √

Smack 0 √

3.3 Procedures

The study consisted of three phases. The first phase was a pre-test of twenty

target words. The second phase was the treatment, which consisted of two

experimental reading tasks and two immediate vocabulary tests. The last phase was a

post-test on target words.

3.3.1 Pre-test of Twenty Target Words

In the first phase, the pre-test made up of twenty target words was administered

to the participants for twenty minutes to measure their prior knowledge about the

target words.

3.3.2 Experimental Reading Tasks and Immediate Post-tests on Vocabulary Gain

In the second phase, all of the participants underwent two experimental reading

tasks, in which they had to finish four reading comprehension questions, along with

two vocabulary immediate post-tests on the target words in two subsequent class

periods.

The passages of reading material were selected from CLASSROOM Smart

Vocabulary for Senior High 7000. The book illustrated the usage of vocabulary in the

7000 list through reading passages under different topics. The two selected reading

passages were Traditional Education and Single Mother. The main idea in the first

passage Traditional Education was about the education during Confucius time

(551-479 B.C.) with a length of two hundred and seven words, and the second

passage was about life of a single mother with a length of two hundred and eight

words.

In the passages, there were sixteen words glossed, inclusive of ten target words

and six non-target words. The presence of the six non-target words was to distract

participants from intentional focus on the target words. To avoid the reading material

from being too long in length, the spacing of the glossed words had to be ignored, so

they were quite close to each other in the reading text.

There were two kinds of gloss types provided for each passage, that is,

meaning-given gloss in the form of single L1 translation and multiple-choice

meaning-inferred gloss in the form of three L1 alternatives. In other words, there were

two passages enhanced with two different versions of gloss types, meaning-given

gloss and meaning-inferred gloss.

One class from the first grade and one from the third grade were assigned to

meaning-inferred multiple-choice group (MI) and the other two classes, including one

from the first grade and one from the third grade, were assigned to meaning-given

group(MG). Participants in the MI group had to finish the multiple-choice gloss by

making the choices of word meanings from three L1 alternatives based on the reading

context. The three L1 alternatives in the meaning-inferred gloss had the same

grammatical category and had similar semantic property. For example, the three

alternatives provided in the glossed word “ emphatic”are all adjectives and each

denotes a type of emotion. Besides, all the L1 alternatives can make sense in the

sentence context, so that it prevents the correct answer from being too obvious.

All participants in the four classes finished four reading comprehension questions

subsequent to the reading material on main ideas or specific details to ensure their

engagement in the meaning-focused reading process (see Appendix B). By providing

reading comprehension questions that participants could review as they read, I

expected that the participants would try to understand the reading passage as a whole

and thereby would not only focus on the target words. To avoid the exposure

frequency effect that the comprehension questions might have on performance of

vocabulary test, target words did not occur in the comprehension questions.

After finishing the reading comprehension questions, participants in the

meaning-inferred (MI) group were offered the correct answers of the multiple-choice

gloss. Then all of the participants in the four groups received the vocabulary

immediate post-test on target words.

3.3.3 Delayed Post-test on Retention

In the third phase, a post-test was administered to measure the retention of the

target words two weeks after the reading tasks. The test items were the same as those

in the pre-test, but they were presented in different order. Since the aim of the study

was to measure the effect of incidental vocabulary learning, the participants were not

informed in advance.

3.4 Data Collection and Analysis

After the three phases of the study, the data was collected and processed in the

following ways. First, to assess the p a r t i c i pa nt s ’word gain, comparisons of the scores

were made between the immediate post-tests and the pre-test. Then, to measure

participants’ word retention, the results of delayed post-test were compared with those

of the immediate post-tests. A series of t-tests were conducted to address the raised

questions in the study.

3.4.1 Independent Variables

Gloss types. The target words in the two reading passages were glossed in two

different conditions, meaning-given gloss and meaning-inferred multiple-choice gloss.

In meaning-given gloss condition, there were sixteen words, including ten target

words and six non-target words glossed in the form of L1 translation immediately

following the target words. As for the meaning-inferred gloss condition, there were

sixteen words, inclusive of ten target words and six non-target words glossed in the

form of multiple-choice comprised of three L1 alternatives for the participants to

choose from and to decide on the meanings of the target words according to the

context (see Appendix B).

Proficiency. In the study, there were two grades of participants, two classes

from the third grade, constituting the groups of higher proficiency level and two

classes from the first grade, making up the groups of lower proficiency level.

3.4.2 Dependent Variables

Vocabulary gain and retention. The p a r t i c i pa nt s ’knowledge of the twenty target

words in the study was measured by the vocabulary pretest, the vocabulary immediate

post-tests after the reading tasks, and the delayed post-test. Then the participants’

vocabulary gain was obtained by a comparison made between their scores of the

immediate post-tests with those of the pre-test. Their vocabulary retention was

measured by means of the comparison made between the scores of the delayed

post-test and those of the pre-tests.

3.4.3 Data Analysis

In order to explore the results of the study, I employed the statistical tool in the

Statistical Package for Social Science (SPSS) to analyze the collected data. The scores

of the vocabulary pre-test, the vocabulary immediate post-tests and the delayed

post-test were compared and analyzed by a series of t-test; the difference between the

two groups (MI and MG) and the effects of different proficiency levels was

investigated.

The first question was answered by comparing the scores of the vocabulary

pre-test with the scores of the immediate post-test through an independent – samples

t-test to see if there was vocabulary gain; the scores of the vocabulary delayed

post-test and the scores of the pre-test was also compared to see if there was word

retention.

The second question was answered by comparing the scores of the vocabulary

gain and retention of MI group with those of MG group through t-test to find out if

there was significant difference between MI group and MG group. In this way,

whether meaning-inferred gloss could generate more incidental vocabulary learning

was tapped.

The third question was answered by comparing the scores of the vocabulary gain

and retention between groups of high proficiency and groups of low proficiency under

two gloss types; t-test was conducted to find out whether there was significant

difference on the comparisons between MI groups and MG groups. In this way,

whether proficiency level affected vocabulary learning under two different gloss types

was investigated.

3.5 Hypothesis

The expected findings for the study were as follows.

1. Meaning-inferred gloss is helpful in incidental vocabulary learning because

of more processing arising from more involvement load. Students'

involvement is enhanced while they are engaged in the multiple-choice

gloss, in which they have the need to search for the appropriate meaning

and evaluate the meaning of the target words in the reading.

2. Learners of high proficiency level will benefit more from meaning-inferred

marginal gloss than learners of low proficiency level because of higher rate

of successful inference resulting from their sufficient knowledge of the

context words and better inferring skills.