To compare the set of transformational leader behaviors identified by Bass
(1985) with those of Kouzes & Posner (1987); exploring potential overlap, superiority, and
The study took place within one of Australia's largest financial
institutions (ANZ Bank); involving 693 branch managers (78% response rate), 1539
subordinates (54% response rate) and 66 District Managers (100% response rate). Twenty
percent of the branch manager's were female. The average age of branch managers was 41
years (ranging from 23 to 58), mean length of time with the bank was 22 years, and average
staff size was 11 people. Eighteen percent had completed or partially completed tertiary
studies (undergraduate college). The majority of subordinates were female (69%). Their
average age was 31 years; they had been with the bank for about 9 years and worked, on
average, with their current branch manager 1.7 years. District managers supervised an
average of 16 branches, were predominantly male (95%), about 46 years old, having worked
for the bank 26 years while in their current positions only 3 years. Just over one-quarter
had completed tertiary studies.
The branch manager's transformational leadership and work performance were rated
by the manager, his/her subordinates, and his/her superior (District Manager). Branch
group cohesion and team performance were rated by the manager and subordinates.
Subordinates' work performance was assessed by the manager and the subordinates.
Managers rated their own self efficacy, and subordinates rated their motivation, extent of
trust in their leader and their leaders' effectiveness.
In addition to the Leadership Practices Inventory (LPI), respondents completed the
Multifactor Leadership Questionnaire Form 5X (MLQ; Bass, 1985) and Global
Transformational Leadership Questionnaire; the latter eight-item scale was developed by
the author. The author developed a Global Work Performance scale, following pilot testing,
resulting in a two-item scale assessing the quality of work and standard of work. Alpha
coefficients of internal reliability were .79 for managers (self) and .82 for subordinates (self);
while for others the alpha ratings were .88 for managers, .88 for subordinates, and .86 for
District Managers. Self-efficacy was adapted from Shearer & Adams (1983) using five
items (alpha = .80). Subordinate motivation (extra effort), modified from several existing
scales, involved four items (alpha = .86). A five-item scale (Ragins, 1989) measured
subordinate perceptions of their leader's effectiveness (alpha = .90). Subordinate trust in the leader used four items from Tjosvold, Andrews & Struthers (1991) with a .90 alpha
coefficient. An eight-item scale, developed for this study, measured team cohesion (alpha
= .89). Six items formed the team performance scale (alpha =.87).
Following several multivariate analyses the author concludes that while
"at the first-order level, in comparison to a one factor model of transformational leadership,
the theoretical models argued by Bass and Kouzes & Posner were found to significantly
fit the data better in both the self and other ratings. However, the fit statistics indicated
that a priori models provided an inadequate account of the observed data. The findings
indicated a 7-factor model for the LPI and a 5-factor model for the MLQ provided a better
fit" (. 119). "Self and other ratings of transformational leadership agree on the same factor
structure" (p. 123).
Using revised LPI and MLQ measures, the author concluded that "there is no
empirical difference between the transformational leader behaviors identified by Bass and
Kouzes and Posner" (p. 151); the LPI is more strongly associated with team performance,
the MLQ more strongly related to individual motivation, and their relationship to leader
effectiveness relatively equivalent.
Combining the LPI and MLQ into a single measure of transformational leadership,
showed that "leadership has a significant impact on team performance..that leader self
efficacy is a significant predictor of transformational leadership and leader work
performance" (p. 176).
Comparisons between self and other ratings, using modified LPI and MLQ scales,
showed managers' self ratings to be consistently higher than those provided by their
subordinates. Further analyses revealed that the strongest correlations between leadership
and work performance is between subordinates, the least agreement between managers and
their subordinates, and some agreement between District Managers (superiors) and
subordinates. Finally, "there is no substantial evidence to indicate females and males differ
in their use of transformational leader behavior" (p. 192).