never answered satisfactorily (Judge et al., 2004). The scales most frequently used to measure the
task- and people-orientation constructs were the initiating structure and consideration scales
developed by the Ohio State investigators, included in the Leader Behavior Description Questionnaire
(LBDQ: (Halpin & Winer, 1957). As Fleishman (1995) QRWHG³&RQVLGHUDWLRQDQGLQLWLDWLQJVWUXFWXUH
KDYHSURYHQWREHDPRQJWKHPRVWUREXVWRIOHDGHUVKLSFRQFHSWV ́(p.51).
%ULHIO\WKH/%'4¶VFUHDWLYHSURFHVVFRQVLVWHGRIWKHUHVHDUFKHUVVSHFLI\LQJQLQHa priori dimensions
of leader behavior. Experts formulated 1,800 items which suited the nine dimensions. Subsequently
this initial set of items was reduced to a set of 150 items. Next, a factor analysis was conducted, using
the Wherry-Gaylord Iterative procedure, on the basis of a sample of 300 members of the Air force
crew who described the behavior of their supervisor. The factor analysis resulted in two factors that
H[SODLQHG RI WKH YDULDQFH DQG WKHVH WZR IDFWRUV ZHUH FDOOHG ³,QLWLDWLQJ VWUXFWXUH ́ DQG
³&RQVLGHUDWLRQ ́ /DWHU VWXGies confirmed these two factors (Fleishman, 1953b; Halpin & Winer,
1957).
Results of the Ohio factor analyses for both the Leader Behavior Description Questionnaire (LBDQ)
(Fleishman, 1953a; Halpin & Winer, 1957) and the Leadership Opinion Questionnaire (LOQ)
(Fleishman, 1953a) indicated that initiating structure (i.e. task-oriented leadership) and consideration
(i.e. people-oriented leadership) were orthogonal dimensions. This orthogonality has been interpreted
as VXJJHVWLQJWKDWWKHWZRGLPHQVLRQVDUHLQGHSHQGHQWWKXVDOHDGHU¶VDFWLRQVRQWKHRQHGLPHQVLRQ
will not affect his/her actions on the other dimension. This orthogonality has been among others
questioned by Bales (1958), Fiedler (1964) and Lowin et al. (1969). The appearance of numerous
sizable correlations in the literature led these authors to seriously question the generality of the
assumed independence of these two leadership dimensions. Weissenberg and Kavanagh (1972) came
to a similar conclusion after reviewing 72 studies. Results from their literature review showed that
51% of the studies reported significant positive correlations, 10% reported significant negative
correlations, and 39% reported non-significant correlations between the dimensions of initiating
structure and consideration.
The type of questionnaire used to measure the dimensions of task-oriented versus relationship-
oriented leadership appears to make a statistical difference. Studies using the Leadership Opinion
Questionnaire (LOQ) reported three positive significant correlations between initiating structure and
consideration, five negative significant correlations, and 16 non-significant relationships. In contrast,
studies using the Leader Behavior Description Questionnaire (LBDQ) reported 34 positive significant