g recognition test in which the target words were presented to learners. The learners were asked to choose their Persian translation.
5. A production test in which some parts of the reading text is presented to learners with target words deleted from the text. The learners were asked to write the word in the blanks.
3.2.2.2. Reading material and target words
For the reading comprehension treatment, two passages from EFL books were chosen. One of them was a passage from “select readings” book with 400 words and the other one a passage from “English result” book consisting of 450 words. These passages were selected randomly from these two books and they include a number of words which were unknown to learners.
3.3. Data collection procedure
A homogeneity test (see Appendix A) was administered to the participants with the purpose of coming up with a homogeneous sample two weeks before treatment. After correcting the papers standard deviation was estimated and the result showed that the selected sample is homogeneous.
One week later, a pre-test (see Appendix B) was administered to ensure that the participants do not know the meaning of the target words. This test was consisted of 25 items including 20 target words that had been chosen from the two texts and five distractors. Each item consisted of a sentence with an underlined new target word, and two choices of ‘Yes’ and ‘No’. Learners were asked to read the sentences and mark ‘Yes’ if they know the meaning of the underlined word and try to provide an English definition or Persian translation and mark ‘No’ if they don’t know the meaning of the underlined words. The participants were given 20 minutes to complete the pre-test. After 20 minutes the pre-test papers were collected and the researcher corrected them. The result showed that no one knows any new target words.
Two suitable reading texts (see Appendix C) from two EFL books were selected. Each one was consisted of 10 new target words. The target words were bolded to raise the learners’ attention. The Persian translations of the words were provided in a parenthesis next to the target words.
After determining the reading texts, the experiment started and conducted within two 120 minutes class period. Two days after pre-test, the first treatment session started. The reading texts were printed for each learner. In a regular day in a class they were asked to read the first text silently. After reading the passage, the participants were asked two questions: a general question about the passage content, and a more specific question which focused on more detailed information. Correct answers to these questions were taken as an indication that they had paid attention to the input. After that, the researcher started to read the text, explain and paraphrase it sentence by sentence and also explained the new target words effectively and provided the synonyms and antonyms for all of them. After finishing this step, the learners were supposed to answer some comprehension questions to be sure that they have comprehended the text.
Two days after the first treatment session, the second session stared. At the beginning of the class the first text was reviewed and then the second text was given to the learners and all the steps in the first treatment session for teaching reading were presented.
After finishing the second treatment session, the participants completed three vocabulary tests (see Appendix D) immediately (N= 60). Each test had 20 questions. The order of the tests was: 1) form recognition test, 2) meaning recognition test, and 3) production test.
3.3.1 Form recognition test
In form recognition test, Persian definitions of target words were provided to the learners and they were asked to choose the target words among options. This test was administered first and lasted for 10 minutes.
The reliability value for this test that was measured by using KR-21 was .95.
The reliability was estimated based on post-test scores.
3.3.2. Meaning recognition test
In meaning recognition test, the target words were presented to learners and they were asked to choose their Persian translation. It was administered right after the learners finished the form recognition test and lasted for 10 minutes.
The reliability value for this test that was measured by using KR-21 was .93.
The reliability was estimated based on post-test scores.
3.3.3. Production test
In production test, some parts of the reading text were presented to learners with target words deleted from the text and the learners were asked to write the word in the blanks. This test was administered after finishing the meaning recognition test and lasted for 20 minutes.
The reliability value for this test that was measured by using KR-21 was .91.
The reliability was estimated based on post-test scores.
It is worth noting that at the beginning of each test, clear instructions were given both orally and in written form to clarify what they were being asked to do. After collecting the papers the researcher corrected the papers and rated all the papers twice.
After two weeks, the participants were asked to take these three vocabulary tests as a delayed post-test (see Appendix E). The delayed post-test was the parallel form of the post-test that was prepared to measure the long term effects of the treatment sessions and at the same time reduce any practice effect. It is worth noting that, all the testing and treatment took place in the classroom setting.
3.4. Data analysis
After collecting the data, correcting the papers, and giving scores, the scores of the vocabulary tests were submitted to SPSS software (version 16) for analyses and descriptive statistics such as mean, standard deviation, and percentage were calculated.
Moreover, inferential statistics such as one-way repeated measure ANOVA was employed. Learners’ scores in three tests (form recognition, meaning recognition, and production) compared via ANOVA to see what aspect of vocabulary knowledge most benefited through reading comprehension.
CHAPTER FOUR
RESULTS AND DISCUSSION
4.0. Overview
In this chapter, the results of data analysis are presented and discussed. Tables are also used for better illustration of the findings. The statistical analysis employed in this study was one-way repeated measure ANOVA.
4.1. Post-test results
4.1.1. Descriptive statistics
Table 4.1 indicates the group performance in the post-test in terms of mean, standard deviation and percentage for the three tests.
Table 4.1. Descriptive statistics of post-test scores
Mean
Std. Deviation
percentage
N
Post-test FR
Post-test MR
Post-test P
17.7750
18.6750
14.3250
2.46501
1.65464
2.94729
.88
.93
.71
40
40
40
FR Form Recognition
MR Meaning Recognition
P Production
As Table 4.1 shows, learners’ mean score for the form recognition test in the immediate post-test is 17.77. Furthermore, the mean score for the meaning recognition test is 18.67, and finally for the production test is 14.32. It also shows that mean score on the meaning recognition test is more than the form recognition and production test.
4.1.2. Inferential statistics
One-way repeated measure ANOVA was used to investigate the difference among learners’ performance in the three tests in the post-test.
Table 4.2 Repeated measure ANOVA result for post-test
Effect
Value
F
Hypothesis df
Error df
Sig.
Partial Eta Squared
Post-test
Pillai’s Trace
.643
34.206
2.000
38.000
.000
.643
Wilks’ Lambda
.357
34.206
2.000
38.000
.000
.643
Hotelling’s Trace
1.800
34.206
2.000
38.000
.000
.643
Roy’s Largest Root
1.800
34.206
2.000
38.000
.000
.643
As Table 4.2 shows, in this study the F value for Wilks’ Lambda is 34.20, with a P value of .000 (which really means p 0.05). The p value is less than 0.05; therefore we can conclude that there is a statistically significant difference among the three sets of scores.
Since one-way repeated measures ANOVA indicated that reading comprehension differentially benefited learners’ three dimensions of vocabulary knowledge, it is now important to see which dimension of vocabulary knowledge benefited more than other dimensions from reading comprehension. To this end, Bonferroni’s pairwise comparisons were performed to locate the differences. Table 4.3 displays the results.
Table 4.3 Pairwise Comparisons result of post-test scores
(I) post-test
(J) post-test
Mean Difference (I-J)
Std. Error
Sig.
95% Confidence Interval for Difference
Lower Bound
Upper Bound
FR
MR
-.900*
.318
.022
-1.695
-.105
P
3.575*
.548
.000
2.205
4.945
MR
FR
.900*
.318
.022
.105
1.695
P
4.475*
.536
.000
3.134
5.816
P
FR
-3.575*
.548
.000
-4.945
-2.205
MR
-4.475*
.536
.000
-5.816
-3.134
FR Form Recognition
MR Meaning Recognition
P Production
This Table compares the learners’ performance in three tests in post-test. It indicates that the learners’ performance in the post-test in the meaning recognition test was statistically higher than learners’ performance in both form recognition and production tests (P 0.05).
It also revealed that the learners’ performance in the post-test in the form recognition test was statistically higher than learners’ performance in the production test.
Thus, the results indicated that reading comprehension has statistically significant effects on three dimensions of vocabulary knowledge. Furthermore, the results indicated that learners’ meaning recognition knowledge of vocabulary was promoted more than the other two aspects by reading comprehension.
4.2. Delayed post-test results
4.2.1. Descriptive statistics
Table 4.1 indicates the group performance in the delayed post-test in terms of mean, standard deviation and percentage for the three tests.
Table 4.4 Descriptive statistics of delayed post-test
Mean
Std. Deviation
percentage
N
Delayed post-test FR
Delayed post-test MR
Delayed post-test P
18.2000
17.0500
13.9750
1.89737
2.46982
2.99989
.91
.85
.69
40
40
40
FR Form recognition
MR Meaning Recognition
P Production
As Table 4.4 shows, learners’ mean score for the form recognition test in the delayed post-test is 18.20, for the meaning recognition test is 17.05 and for the production test is 13.97. Furthermore, as the Table shows, the mean score on the form recognition test is more than the meaning recognition and production test.
4.2.2. Inferential statistics
One-way repeated measure ANOVA was used to investigate the difference among learners’ performance in the three tests in the delayed post-test.
Table 4.5 Repeated measure ANOVA result for delayed post-test
Effect
Value
F
Hypothesis df
Error df
Sig.
Partial Eta Squared
Delayedpost-test
Pillai’s Trace
.620
31.062
2.000
38.000
.000
.620
Wilks’ Lambda
.380
31.062
2.000
38.000
.000
.620
Hotelling’s Trace
1.635
31.062
2.000
38.000
.000
.620
Roy’s Largest Root
1.635
31.062
2.000
38.000
.000
.620
Table 4.5 reveales that the F value for Wilks’ Lambda is 31.06, with a P value of .000. The p value is less than 0.05 (P 0.05); thus it indicates that there is a statistically significant difference among the three sets of scores in delayed post-test.
since one-way repeated measures ANOVA revealed that reading comprehension has an effect on three dimensions of vocabulary knowledge, it is important to know which dimension of vocabulary knowledge benefited more than other dimensions from reading comprehension after a long time. Therefore, Bonferoni’s pairwise comparisons used to locate the differences among the three tests. Table 4.6 shows the results. It compares the learners’ performance in three tests in delayed
