Results
Students’
Concept Maps
The difference in concept
maps between computer-based and paper-and-pencil concept mapping was
examined based on two concept map measures, the number of ideas contained
in the maps and the total quality of the maps, using the nested factorial
design described previously. Table 1 presents the means and standard
deviations for these measures by method, teacher, and class.
Table 1.
Means and Standard Deviations for Concept Mapping Measures,
and Persuasive Writing Postassessment Scores by Method by Teacher by
Class
|
|
| Measure |
Class |
n |
Number
of Ideas
|
Total
Quality
|
Writing |
|
|
| |
|
|
M |
SD |
M |
SD |
M |
SD |
|
|
| Paper-and-Pencil |
|
|
|
|
|
|
|
|
| Teacher
1 |
1 |
12 |
12.92 |
2.71 |
24.00 |
8.68 |
3.50 |
.52 |
| |
2 |
15 |
12.60 |
1.84 |
25.33 |
6.30 |
3.07 |
.80 |
| Total |
3 |
27 |
12.74 |
2.23 |
24.74 |
7.33 |
3.26 |
.71 |
| Teacher
2 |
1 |
16 |
10.44 |
4.83 |
17.50 |
15.28 |
3.13 |
.81 |
| |
2 |
15 |
9.20 |
4.16 |
16.53 |
12.97 |
3.13 |
.92 |
| |
3 |
15 |
8.60 |
3.07 |
16.67 |
10.91 |
3.33 |
.62 |
| |
4 |
18 |
8.67 |
3.46 |
17.17 |
11.40 |
3.89 |
.90 |
| |
5 |
13 |
10.23 |
3.77 |
19.38 |
10.84 |
3.46 |
.52 |
| Total |
5 |
77 |
9.39 |
3.88 |
17.39 |
12.15 |
3.40 |
.82 |
| Total |
7 |
104 |
10.26 |
3.81 |
19.30 |
11.53 |
3.37 |
.79 |
|
|
| Computer-based |
|
|
|
|
|
|
|
|
| Teacher
1 |
1 |
16 |
9.13 |
3.65 |
15.00 |
9.63 |
3.19 |
.91 |
| |
2 |
15 |
10.00 |
2.42 |
20.20 |
7.77 |
2.87 |
.64 |
| Total |
3 |
31 |
9.55 |
3.10 |
17.52 |
9.03 |
3.03 |
.80 |
| Teacher
2 |
1 |
19 |
10.68 |
5.41 |
24.21 |
20.83 |
2.68 |
.82 |
| |
2 |
17 |
11.59 |
5.54 |
17.59 |
11.52 |
3.29 |
.59 |
| |
3 |
20 |
15.85 |
5.90 |
34.60 |
17.37 |
2.95 |
.83 |
| |
4 |
18 |
12.94 |
5.76 |
30.06 |
15.36 |
2.89 |
.90 |
| |
5 |
17 |
14.76 |
3.70 |
32.71 |
14.39 |
2.88 |
.70 |
| Total |
5 |
91 |
13.20 |
5.40 |
28.00 |
17.14 |
2.93 |
.79 |
| Total |
7 |
122 |
12.27 |
5.16 |
25.34 |
16.11 |
2.96 |
.79 |
|
|
The Method III sums of squares
ANOVA for the numbers of ideas contained in the students’
concept maps revealed a statistically significant effect for concept
mapping method, F (2, 1.94) = 42.02, MSE = 286.21,
p = .03, 
= .15. A
statistically significant effect was also found for teacher within method,
F (2, 10.79) = 6.78, MSE = 38.26, p = .01,
= .06. The effect for classes
within teacher within method was also statistically significant, F
(10, 212) = 2.24, MSE = 17.87, p = .02,
= .11. The students who developed computer-based concept maps (M
= 12.27) generated more ideas (p < .05) than the students
who developed paper-and-pencil concept maps (M = 10.26). This
means the computer-based concept mapping facilitated the brainstorming
phase of prewriting and idea generation for a persuasive writing task.
In addition, the results indicate that teachers influenced the number
of ideas generated by their students regardless of the method of concept
mapping (
= 3.92). The number
of ideas generated also varied by class (
= 1.38).
The Method III sums of squares
ANOVA results for the total quality of the students’
concept maps displayed a statistically significant effect for concept
mapping method, F (2, 1.92) = 24.53, MSE = 1888.18,
p = .04, = .10, a
statistically significant effect for teacher within method, F
(2, 10.85) = 4.82, MSE = 357.59, p = .03,
= .04, and a statistically significant effect for class within teacher
within method, F (10, 212) = 2.10, MSE = 178.11, p
= .03, = .10. The students
in the computer-based concept mapping condition (M = 25.34)
produced higher total quality concept maps (p < .05) than
the students in the paper-and-pencil concept mapping condition (M
= 19.30). Similar to the results for the number of ideas contained in
the students’ concept maps, the total quality findings indicate
that teachers influenced the quality of students’ concept maps
no matter the method of mapping used by the students (
= 24.20) and that the quality of the students’ concept maps varied
by class ( = 12.20).
Post hoc mean comparisons
using the Tukey-Kramer procedure indicated that the mean performance
(M = 28.00) level of the students’ with the second teacher
in the computer-based mapping condition differed (p < .05)
from the mean performance
(M = 17.39) level of the students with the second teacher in
the paper-and-pencil concept mapping condition. None of the other differences
among the teachers were statistically significant. Inspection of the
students’ concept maps revealed that the students with the second
teacher in the computer-based concept mapping condition closely adhered
to the persuasive concept map template while the students with the second
teacher in the paper-and-pencil concept mapping condition did not adhere
as closely to the persuasive concept map template. Hence, the total
quality scores were partly affected by adherence to the persuasive concept
map template and the teachers made a difference as to whether the students
followed the template.
Effect of Concept
Mapping Method on Persuasive Writing Performance
The effect of concept mapping
prewriting method (computer-based versus paper-and-pencil) on eighth-grade
language arts students’ persuasive writing was assessed using
ANCOVA. The students’ scores on the persuasive writing preassessment
served as the covariate. Table 1 presents the means and standard deviations
for the two concept mapping methods on the students’ persuasive
writing postassessment scores by method, by teacher, and by class.
Results of the Method III
sums of squares ANCOVA showed a statistically significant effect for
the prewriting covariate, F (1, 211) = 108.19, MSE
= .395, p < .01. The effect of mapping method was statistically
significant, F (2, 5.14) = 47.08, MSE = .363, p
< .01, = .17. The effect of teacher within method
was not statistically significant, F(2, 10.90) = .52, MSE
= .72, p = .61, =
.00. The effect of class within teacher and method was statistically
significant, F(10, 211) = 1.88, MSE = .40, p
= .05, = .02. This means the
students’ persuasive writing scores varied by class (
= .02). Contrary to our expectations, the students who developed their
concepts maps using paper-and-pencil (M = 3.37) performed better
than the students who developed their concepts maps using the computer
software (M = 2.96).
Relation of Concept
Map Quality to Persuasive Writing
Stepwise multiple regression
analysis was used to evaluate the relation of the concept mapping measures
to persuasive writing. Table 2 displays the correlations among these
measures. The stepwise regression was statistically significant, F
= 7.14, MSE = .64, p < .01, R2
= .03. Partially due to a high correlation between the total quality
of the students concept maps and the number of ideas contained in their
maps (r = .83), the total quality of the students’ concept
maps was selected as the only predictor of the students’ persuasive
writing scores. However, the quality of the students’ concept
maps accounted for only three percent of the differences in the quality
of the students’ persuasive writing.
Table 2.
Correlation of Concept Mapping Scores with Persuasive
Writing Postassessment Scores (n = 226)
|
| Variable |
Total
Quality |
Postassessment
Writing |
|
| Number
of Ideas |
.83** |
.12* |
| Total
Quality |
|
.18** |
|
| *p
< .05. **p < .01 |
|
|
