Effective reading comprehensio
Effective reading comprehensioEFFECTING READING COMPREHENSION WITH COMPUTER QUIZZES
Abstract
The purpose of this study was to investigate the effect of periodic computer-based comprehension quizzes in reading classrooms. Two special education resource-reading classes were used in this study. The students in one of the classes (Class X) received periodic computer-based comprehension quizzes as they read various books. The students in the other class (Class Y) did not. Both classes received computer-based comprehension tests upon completion of reading various books. After 5 six week periods (5/6 of the school year), the overall average grades were compared of 10 students randomly selected from each of the two classes. The results of the study did not show any conclusive evidence that the use of computer-based comprehension quizzes improved overall reading comprehension performance.
The Effects of Periodic Computer -Based Comprehension Quizzes in the Reading Classroom
Background Information
This study assessed differences among two special education resource-reading classes. The classes were two separate resource-reading classes at Flintville Junior High School in Flintville, Tennessee. Two separate teachers taught the two classes. During the 1998-1999 school year one of the resource teachers used computers in his classroom to give periodic comprehension quizzes to the reading students as they read various stories. The second teacher did not use computers for periodic comprehension quizzes.
At the end of the school year, the grades of the students for the year were compared. The purpose of the study was to determine whether there appeared to be any significant difference in the overall progress made in reading comprehension between the two classes.
Problem Statement
To establish a main focus in the study, the following problem statement was used: What are the differences in the overall reading comprehension grades of two separate resource-reading classes if one class is periodically quizzed with computer tests and the other is not?
Hypothesis Statement
If two separate resource-reading classes are tested on reading comprehension upon completion of stories they read, and one of the classes is given computer-based comprehension quizzes periodically as they read the stories, then there will be a difference in the overall reading comprehension grades of the two classes.
Hypothesis Rationale
Although only one of the classes was periodically given computer-based comprehension quizzes during the time they read stories, both resource-reading classes were evaluated upon completion of stories by Accelerated Reader computer-based tests. The Accelerated Reader program is a computer software program that uses computer-based multiple-choice tests to measure student's reading comprehension. The better students score on the tests, the more "points" they acquire. These points may later be used by students to "buy" incentive prizes such as yo-yo's, sunglasses, and other various prizes offered by the school librarian. It is an example of how computers are now being used in all subject areas in schools. In the past, computers were thought to only be useful in academic areas such as science and math. Recently, however, computers are now being used in other areas such as social studies, reading, writing, and foreign languages (Fresch, 1999). Students using computers in language arts classes have demonstrated in one study that their incentive to read increases and they read more carefully (Sullivan, 1998).
There are two sides to the issue of using computers in schools. On the flip side of the "computer revolution" in our schools are those who feel that, although computers can be useful tools, they are not the end in themselves (Ravich, 1998). The author of a recent article in Time magazine stated that students should first learn reading and writing, history and arithmetic. Then "surfing the internet" or whatever other "fun" things could follow (Gelernter, 1998). Another author recently stated in Teaching PreK-8 magazine, that technology isn't a substitute for traditional basics. It's just one part of a comprehensive education agenda (Riley, 1998). Whether either side of this debate (or both) are right remains to be seen.
As the school year progressed, the teachers in our study observed that their classes seemed to be progressing at different rates. Since the class that was given periodic computer-based comprehension quizzes seemed to be accumulating Accelerated Reader points at a faster rate, the inductive hypothesis emerged that the periodic quizzes may be at least partially responsible. Data was collected which seemed to substantiate the inductive hypothesis with deductive evidence. That is, the class that was given periodic computer-based comprehension quizzes was, in fact, accumulating Accelerated Reader points at a faster rate.
Population Description
The defined population was made up of students from two resource-reading classes. Table 1 shows percentage scores of Accelerated Reader tests taken by the students in the two resource-reading classes. The students who were given the additional periodic computer-based reading comprehension quizzes were in "Class X". The students who were not given the additional periodic computer-based reading comprehension quizzes were in "Class Y".
All of the students used in this study had evaluated reading levels ranging from 1st grade level to 4th grade level. The students IQ's ranged from the mildly mentally retarded level (MMR) to the learning disabled (LD) level. These two levels were not equally distributed. Class X had a concentration of (MMR) students and Class Y had a concentration of (LD) students. However, since all of the students in this study were reading books on their appropriate reading levels, it was not felt that this would have an adverse effect on the results of the study.
Sampling Method
The method of sampling used in this study was stratified sampling. 10 students were randomly selected from each class by choosing every other one that appeared alphabetically in the teacher's roll books. Therefore, N=10 for both classes in the study.
Instrument Descriptions
The data was collected from the student's cumulative averages through the 5 six- weeks period. There are 6 six-week periods in the school year so this cumulative average represented 5/6 of the students average for the school year. The independent variable in this study was the periodic computer-based reading comprehension quiz administered to Class X as they read each story. These tests were written and administered by the teacher of Class X using the "Half-Baked Hot Potatoes" software that was developed at the University of Victoria in Canada. The dependent variable in this study was the cumulative reading comprehension averages at the end of the 5th six -weeks period. Averaging the Accelerated Reading comprehension test scores derived these averages. The Accelerated Reading comprehension tests were given to all of the students upon completion of reading stories.
Validity of Instruments
According to the manual that accompanies the Accelerated Reader kit, the validity of the test is .63. The Accelerated Reader program is widely used in many school systems today. One of the main reasons for this is the content validity of the tests, which consist of the item validity and the sampling validity.
The item validity is more than adequate because every item on each test deals specifically with the particular book that the student is being tested on. In other words, if a student is being tested on The Adventures of Huckleberry Finn, he or she will not encounter any questions about the author (Mark Twain) or other books written by him, (i.e., The Adventures of Tom Sawyer, etc.).
Likewise, the sampling validity is satisfactory because the test questions are extracted proportionally from the entire book and not from concentrated areas. Therefore, the student must have read the entire book in order to be able to correctly answer all of the comprehension questions.
Since the periodic comprehension quizzes given to Class X were constructed using the same (or very similar) questions as the Accelerated Reader questions, the validity of both test methods are identical.
Reliability of Instruments
According to the manual that accompanies the Accelerated Reader kit, the reliability of the test is in the range of the upper .80's to lower .90's, depending on the grade equivalency or reading level being tested. Again, since the periodic comprehension quizzes used with Class X were constructed using the Accelerated Reader questions, the reliability of both tests would also be identical.
For this study, the rationale equivalence reliability of the periodic computer-based comprehension quizzes (and the Accelerated Reader tests) was estimated by using the Kuder-Richardson formula 21 (KR-21). An example of this method follows using a pilot sample of tests results from a class of 10 students taking a 20 item comprehension test:
Pilot Sample Grade Averages
75
80
90
65
70
75
80
85
90
100
From this sample, we discovered that our mean () = 81 and the standard deviation (SD) = 9.95. Using this information with the KR-21 formula:
r total test = (K) (SD�) - (K- )
(SD�) (K - 1)
K = the number of items on the test
SD = the standard deviation of the scores
= the mean of the scores
We then calculated as follows:
r total test = (20) (99.0025) - 81 (20- 81)
(99.0025) (20 - 1)
r total test = 1980.05 - 81 (-61)
(99.0025) (19)
r total test = 1838.05
1881.05 = .98
Therefore, the rationale equivalence reliability of the periodic computer-based comprehension quizzes (and the Accelerated Reader tests) = .98.
Data Presentation and Analysis
As mentioned previously, the rationale equivalence reliability of the periodic computer-based comprehension quizzes (and the Accelerated Reader tests) was estimated by using the Kuder-Richardson formula 21 (KR-21). This time the actual averages of test results from each class of 10 students taking a 20 item comprehension test was used:
Actual Grade Averages
Class X Class Y
80 78
75 72
90 90
90 87
75 85
87 75
88 75
90 75
86 80
85 85
For Class X, we discovered that our mean () = 84.6 and the standard deviation (SD) = 5.59. Using this information with the KR-21 formula:
r total test = (K) (SD�) - (K- )
(SD�) (K - 1)
K = the number of items on the test
SD = the standard deviation of the scores
= the mean of the scores
We then calculated as follows:
r total test = (20) (31.2) - 84.6 (20- 84.6)
(31.2) (20 - 1)
r total test = 624 - 84.6 (-64.6)
(31.2) (19)
r total test = 474.8
592.8 = .80
Therefore, the rationale equivalence reliability of the periodic computer-based comprehension quizzes (and the Accelerated Reader tests) for Class X = .80.
Likewise, For Class Y, we discovered that our mean () = 80.2 and the standard deviation (SD) = 5.84. Using this information with the KR-21 formula we then calculated as follows:
r total test = (K) (SD�) - (K- )
(SD�) (K - 1)
K = the number of items on the test
SD = the standard deviation of the scores
= the mean of the scores
r total test = (20) (34.1) - 80.2(20- 80.2)
(34.1) (20 - 1)
r total test = 682 - 80.2(-60.2)
(34.1) (19)
r total test = 541.6
647.9 = .84
Therefore, the rationale equivalence reliability of the Accelerated Reader tests for Class Y = .84.
Since both of these rationale equivalence reliabilities were acceptable, the next step in the study was to determine if there was a significant difference in the overall reading comprehension averages of the two classes.
The following table shows the overall averages of the student's reading comprehension grades in Class X and Class Y for at the end of the 5th six weeks period (5/6 of the school year):
Table 1
Class X Class Y
80 78
75 72
90 90
90 87
75 85
87 75
88 75
90 75
86 80
85 85
Using the data from Table 1, calculations were made for the independent t test using the following data and independent t formula for raw scores:
X Y X� Y�
80 78 6400 6084
75 72 5625 5184
90 90 8100 8100
90 87 8100 7569
75 85 5625 7225
87 75 7569 5625
88 75 7744 5625
90 75 8100 5625
86 80 7396 6400
85 85 7225 7225
X 846
84.6 Y 802
Y 80.2 X� 71884 Y� 64662
Independent t formula for raw scores:
T= - Y
By using the formula on the preceding page, the independent t value was 1.63. We then determined the degrees of freedom using the following formula:
df=N1 + N2 -2
This calculation resulted in the degrees of freedom being 18. Looking at the t table of critical values (cv) we saw that cv= 2.101 (Gay, 1996). Therefore, our t value was less than our critical value and not large enough to allow us to reject the null hypothesis.
Discussion of Results
After determining that the rationale equivalence reliabilities were acceptable, the results of the study went on to show that the null hypothesis must be accepted. That is, any difference between the overall performance of Class X and Class Y may be entirely due to chance. Although it appeared at various times during the year, that Class X may have been out-performing Class Y, in the long run there was not enough evidence to prove that Class X's periodic computer-based comprehension quizzes were any more effective than Class Y's method of reading instruction.
There is no doubt that computers can be very useful tools in the classroom. It is also apparent that, with 70% of America's classrooms on-line, it is inevitable that every classrooms will one day be on-line (Ravich, 1998). We are still not, however, at the point where it is clear when the computer can help us teach students better than traditional methods. All that we can do is to continue to do research like this study until we know more.
In conclusion, one thing that is certain is the fact that the use of computer-based instruction and testing has increased the level of interest among the students. And all teachers are for anything that does that.
References
Fresch, M.J. (1999, March). Alice in computerland: Using the Internet as a resource for teaching reading. Reading Teacher, 52, 652-653.
Riley, R.W. (1999, October). Technology in every classroom. Teaching PreK-8, 29 (2), 8.
Sullivan, J. (1998, September). The electronic journal: Combining literacy and technology. Reading Teacher, 52 (1), 90-94.
Gelernter, D. (1998, May). No--learn first, surf later. Time, 151(20), 55-58.
Ravich, D. (1998, March). The great technology mania. Forbes, 161 (6), 134