Character for Leadership: The Role of Personal Characteristics

Analyses of Variance 
Analysis of variance (ANOVA) is used to determine if there are differences 
between the means of three or more discrete groupings for a variable. Group 

---

Character for Leadership 
54 
 
members were identified as those whose responses fell within the lowest third, 
middle third, and upper third percentiles, respectively, for each character scale. The 
traditional ANOVA is based on the assumptions that the population from which the 
data are taken is both normally distributed and is homoscedastic. The data in this 
study meet the criterion of equal variances according to Levene’s test. Table 6 
presents this information. 
Table 6: Levene’s Test for Equality of Variances 
Levene statistic 
p 
H
1
2.90 
.06 
H
2
1.19 
.31 
H
3
1.62 
.20 
H
4
.24 
.79 
H
5
.55 
.58 
H
6
1.10 
.34 
H
7
.14 
.87 
However, due to the noncontinuous nature of the independent variable data, 
a nonparametric test must be used to examine differences between means. 
Therefore, the data for this study were analyzed using the nonparametric equivalent 
of the traditional ANOVA, the Kruskal–Wallis test. Data were analyzed for 
hypotheses 1 - 6 by Kruskal–Wallis to determine whether or not there were 
significant differences in character groups (low, medium, and high levels) for each 
leadership scale. Data for hypothesis 7 were also analyzed for differences between 
groups due to the division of data for the independent variable, current ministry 
leadership involvement, into three discrete groups. Kruskal–Wallis was also used to 
analyze data for hypothesis 7. Table 7 presents the results of all data analyses for 
the stated hypotheses. An alpha level of .05 was used for all statistical tests. 

---

Character for Leadership 
55 
 
Table 7: Chi-Squared Results for Hypotheses,
α
= .05 
χ
2
p 
H
1
Confident leadership by self-directedness level 
20.37 
.00
H
2
Visionary 
leadership 
by 
self-directedness level 
17.29 
.00
H
3
Follower-centered leadership by cooperativeness level 
7.97 
.02
H
4
Visionary leadership by cooperativeness level 
18.40 
.00
H
5
Mature character by self-transcendence level 
2.87 
.24
H
6
Visionary leadership by self-transcendence level 
.22 
.95
H
7
Visionary leadership by current ministry level 
5.45 
.07
Hypothesis 1 posited a difference in how the VLT construct of confident 
leadership was enacted between groups based on measured levels for the character 
trait self-directedness. The difference between participants demonstrating low, 
medium, and high levels of self-directedness was found statistically significant for 
confident leadership, 
χ
2
(2, 
Ν
= 101) 
= 20.37, 
p
= .00. This hypothesis was 
supported, allowing rejection of the null hypothesis at that level. Hypothesis 2 
posited a difference in how visionary leadership [the total score on Sashkin et al.’s 
(1997) TLP] was enacted between groups based on measured levels for the 
character trait self-directedness. The difference between participants demonstrating 
low, medium, and high levels of self-directedness was found statistically significant 
for visionary leadership, 
χ
2
(2, 
Ν
= 101) 
= 17.29, 
p
= .00. This hypothesis was also 
supported, allowing rejection of the null hypothesis at that level. Hypothesis 3 
proposed a difference in how the VLT construct of follower-centered leadership 
was enacted between groups based on measured levels for the character trait 
cooperativeness. The difference between participants demonstrating low, medium, 
and high levels of cooperativeness was found statistically significant for follower-
centered leadership, 
χ
2
(2, 
Ν
= 101) 
= 7.97, 
p
= .02. This hypothesis was supported, 
allowing rejection of the null hypothesis at that level. Hypothesis 4 proposed a 
difference in how visionary leadership was enacted between groups based on 
measured levels for the character trait cooperativeness. The difference between 

---

Character for Leadership 
56