7
Table 3. The Features of Experiment 2 and Experiment 3.
| Procedure | Differences | |
| Experiment 2 | Experiment 3 | |
| Pre-test | Yes | Yes |
| Posttest | Yes | Yes |
| Meanings provided after post-test | Yes | No |
| Text Given after post-test | Yes | No |
| Story Used | 3 parts of one story | 3 different stories |
| Number of Words | 103 | 60 |
| Delayed post-test | Yes 4~7 weeks later | Yes 4~7 weeks later |
| Number of Participants | 7 | 6 |
| Teacher | MV | KY |
| Time spent for the lesson | 135 minutes (30 minutes for listening and 15 minutes for corrective feedback and answer check for each lesson) | 60 minutes (20 minutes for story listening only for each lesson) |
Results
The mean pre-test score was 10.7, or 18%, nearly identical to the percentage correct in experiment 2 (19.6/103 = 19%), confirming that the lists were of similar difficulty and that most words were unknown to the subjects.
Table 4 presents descriptive statistics for the pre-, post- and, delayed post-tests. There were 60 words in total. The participants scored 10.7 correct on the pre-test. While they listened to the stories, they learned 28.5 new words, but they eventually forgot an average of 22.5 of them. They thus remembered six words, or 12% of those they learned, after four to seven weeks. The rate of acquisition/learning was .10, six words in 60 minutes, or one-tenth of a word per minute.
Table 4. Mean and S.D. of Pre-, Post-, and Delayed post-test
| Pre-test Mean (S.D.) | Post-test Mean (S.D.) | Gain | Delayed Mean (S.D.) (4~7 wks Later) | Final Gain | Time Spent | Rate Words/min. |
| 10.7 (5.3) | 39.2 (7.3) | 28.5 (58%) | 16.7 (7.2) | 6 (12%) | 60 min. | .10 |
Total number of words on the tests = 60
Unknown words = 60 - 10.7 = 49.3
Posttest = 39.2 correct, but they knew 10.7. 39.2-10.7=28.5. 28.5/49.3 = 58%
Delayed posttest = 16.7 but they knew 10.7. 16.7-10.7=6. 6/49.3 = 12%
There was a statistically significant difference between the pre-test and the delayed post-test scores (t test for correlated samples, t = 3.80, df = 5, p < .01).
Discussion
The figures presented here are undoubtedly an underestimate of language acquisition from the stories: Students may have acquired other words that were not focused on, as well as other aspects of language (grammar and pronunciation). Also, our tests only measured relatively complete vocabulary acquisition, to the point of being able to supply a synonym. It is possible, and likely, that subjects acquired parts of the meanings of unfamiliar words, even if they could not produce a translation (Nagy, Herman and Anderson, 1985).
7