134 PREPARATION OF SCHOOL LEADERS
The qualitative data gathered using the open-ended survey were analyzed by first reading
all responses. Responses were then coded using open coding, as described by Strauss and
Corbin (1998). Responses were then aggregated into categories based on similarity.
RESULTS
Quantitative Data
The mean scores and standard deviations of the control and treatment groups on the exam
are shown in Table 2.
Table 2. Mean Scores and Standard Deviations by Test Format.Group Test format Mean score SD
Control individual test 17.78 2.46
Treatment Group 1 paired test 19.08 0.82
Treatment Group 2 individual test by choice 18.76 1.80
Treatment Group 3 paired test by choice 18.70 1.79
A significant difference was found between the scores of students in the control group and
students in Treatment Group 1. The control group’s mean score was significantly lower than
Treatment Group 1’s mean score, t (83) = 4.62, p = .00. Pairing for testing improved students’
achievement on the exam with an effect size of 0.45.
Also significant was the overall difference between the scores of students who tested indi-
vidually and those who tested in pairs. Allowing students to select their own team was rec-
ommended by 3 of the students in Treatment Group 1 during the first part of this study and is
addressed by both Hancock (2007) and Zimbardo et al. (2003). Zimbardo and colleagues
found that whether self-selected or paired, students performed better when teamed than when
assessed individually. When the control and Treatment Group 2 scores were combined (M =
17.82, SD = 2.44), and when Treatment Groups 1 and 3 were combined (M = 18.88, SD =
1.41), there was a significant difference, t (120) = 3.66, p = .00, with those testing in pairs
outperforming those testing individually. Whether paired or selecting the pair option, teaming
improved students’ achievement with an effect size of 0.53.
The data collected indicate that there were no significant differences between the mean
group scores for Treatment Groups 1 and 2, t (4) = 0.39, p = .72, Treatment Groups 1 and 3, t
(30) = 0.90, p = .38, or Treatment Groups 2 and 3, t (25) = –0.07, p = .95; that is, the average
performance of each treatment group was not significantly different from another. Differences
between the control group and Treatment Group 2, t (136) = –0.88, p = .38, and Treatment
Group 3, t (153) = –1.68, p = .10, were also found to be not significant.
Qualitative Data
Of the 20 students who completed the survey instrument, 18 said that they would like to
team test again. This level of student approval is consistent with data in the literature. Zim-
bardo et al. (2003) found that students overwhelmingly support the use of team testing as an
assessment tool. The reasons students in this sample would like to team test again can be
summarized by the following: