Character for Leadership: The Role of Personal Characteristics

Multiple Regression Analysis 
Though not part of the original hypotheses, because of the exploratory 
nature of the study, the researcher chose to utilize multiple regression analysis to 
explore the degree of variance in visionary leadership that could be attributed to the 
character traits of the leader. Multiple regression analysis was designed to explain 
the relationship between multiple independent variables and one dependent variable 
even though the results, in and of themselves, are not designed to assess causality 

---

Character for Leadership 
57 
 
(Cohen, Cohen, West, & Aiken, 2003). In addition to leader character traits, the 
demographic variables related to ministry leadership experience (previous 
vocational leadership experience and current weekly vocational leadership 
involvement) were also loaded in the regression model to explore the degree of 
variance that could be attributed to them. Though the previous experience variable 
was not specifically identified in the literature as a variable of interest, it is similar 
in nature to that of current ministry leadership involvement and may therefore have 
a similar effect on the dependent variable.
Backward regression analysis was performed in the first analysis to assess 
the amount of variance accounted for in the dependent variable (visionary 
leadership behaviors) by the independent variables (self-directedness, 
cooperativeness, and self-transcendence). In backward regression analysis, all 
independent variables are entered into the regression model and then those that do 
not contribute significantly to the regression equation, based on statistical analysis, 
are removed from the model. 
The results of this analysis demonstrate that of the three character trait 
independent variables, only self-directedness and self-transcendence remained in 
the model. Table 8 shows the combined effects of these two variables account for 
31% of the variance in the dependent variable, visionary leadership behaviors (
R
2
= 
.31), and multiple correlation of .55 of these two variables with the dependent 
variable. 
Table 8: Regression Model – Character Traits 
Model
R
 
R
2
 
F
 
df 
1
df 
2
p
 
1
.56
a
.32
15.07
3
97
.00
2
.55
b
.31
21.61
2
98
.00
a
Predictors: (constant), self-transcendence, self-directedness, cooperativeness. 
b
Predictors: (constant), self-transcendence, self-directedness. 
Furthermore, as shown by Table 9, variance in visionary leadership 
behaviors is primarily attributed to self-directedness (ß = .23 for self-transcendence 

---

Character for Leadership 
58 
 
and visionary leadership which is statistically significant though moderately 
correlated), thus indicating the relative significance of the character trait self-
directedness. 
Table 9: Regression Coefficients – Character Traits, Dependent Variable Visionary 
Leadership
Unstandardized 
coefficients
Standardized coefficients
 
Model
B
 
SE
ß
t
p
1
(Constant)
101.71
21.83
4.66
.00
Self-transcendence
1.40
.59
.21
2.39
.02
Cooperativeness
.90
.70
.13
1.29
.20
Self-directedness
1.32
.30
.45
4.41
.00
2
(Constant)
126.96
9.83
12.92
.00
Self-transcendence
1.58
.57
.23
2.77
.01
Self-directedness
1.54
.25
.53
6.22
.00
Forward stepwise regression analysis was used in a second analysis to 
assess the effects of ministry leadership on visionary leadership behaviors of study 
participants. In stepwise regression, entry of the variables into the regression model 
is done statistically based on those having the highest correlation with the 
dependent variable. Table 10 shows that three variables entered the model: self-
directedness, self-transcendence, and previous leadership experience. The effects of 
these three variables combined explained 33.8% of the variance in the dependent 
variable visionary leadership behaviors (
R
2
= .338). 

---

Character for Leadership 
59 
 
Table 10: Regression Model – Character Traits and Previous Leadership 
Experience 
Model
R
 
R
2
 
F
 
df 
1
df 
2
p
 
1
.502
a
.252
33.33
1
99
.00
2
.553
b
.306
21.61
2
98
.00
3
.582
c
.338
16.53
3
97
.00
a
Predictors: (constant), self-transcendence, self-directedness, cooperativeness. 
b
Predictors: (constant), self-transcendence, self-directedness.
c
Predictors: (constant), self-transcendence, self-directedness, leadership experience. 
Table 11 shows that when controlling for the effects of self-directedness, 
regression coefficients indicate only a moderate though significant correlation (ß = 
.21) between the character trait self-transcendence and visionary leadership 
behaviors and a weak though significant correlation (ß = .18) between previous 
leadership experience and visionary leadership behaviors. These results together 
indicate that while self-transcendence and one’s previous experiences do affect 
one’s ability to behave in a manner consistent with the ideals of visionary 
leadership, the character trait self-directedness explains the most variance (ß = .50) 
in the selection of those behaviors. 

---

Character for Leadership 
60 
 
Table 11: Regression Coefficients – Character Traits and Previous Leadership 
Experience, Dependent Variable Visionary Leadership
Unstandardized 
coefficients
Standardized coefficients
Model
B
 
SE
 
ß
t
p
1
(Constant)
140.02
8.90
15.73
.000
Self-directedness
1.47
.26
.50
5.77
.000
2
(Constant)
126.96
9.83
12.92
.000
Self-directedness
1.54
.25
.53
6.22
.000
Self-transcendence
1.58
.57
.23
2.77
.007
3
(Constant)
124.82
9.70
12.87
.000
Self-directedness
1.46
.25
.50
5.89
.000
Self-transcendence
1.41
.57
.21
2.49
.015
Leadership experience
2.62
1.21
.18
2.17
.032