Data was collected by means of an online questionnaire (survey) asking participants to rate the degree
to which a set of behavioral strategic leadership statements suited the focal PDQDJHU¶VEHKDYLRUThe
behavioral items were scored along a continuous 5-step Likert scale. The number of scale steps for
continuous scales is generally recommended from five to seven (Pett et al., 2003). Fewer choices of
VFDOHVWHSV HJWZRWKUHHRUIRXU UHVWULFWWKHLWHP¶VYDULDQFH(YHQQXmbers of scale steps force the
rater to either agree or disagree. The problem is that the subject may, in reality, be undecided. An odd
number allows for the middle scale step to be the neutral, or indifferent point. The response options
were worded as such that there were approximately equal intervals of agreement between them
³WRWDOO\GLVDJUHH ́³GLVDJUHH ́³LQ-EHWZHHQ ́³DJUHH ́³WRWDOO\DJUHH ́(DeVellis, 1991).
6.2.2 MEASURES
Based on the literature review (chapter 5), 24 items were formulated, which were considered to best fit
the four strategic leadership dimensions. The next step in the scale development consisted of content
expert validation, which is a method to guarantee the content validity of the measurement instrument
(Grant & Davis, 1997). It is essentially used to identify which items are incoherent (Hinkin, 1995).
The Standard for Educational and Psychological Testing (American Educational Research
Association, 1985) hold two criteria to determine who to include in an expert panel, namely relevant
training experience, and qualifications. With these criteria in mind, the content experts were sourced
from top levels in organizations. The sampling technique was purposive and resulted in a selection of
interview participants that were leaders with relevant experience with the research topic (Minichiello
et al., 1995). Five senior executives have been consulted, which made the sample size for the content
analysis relatively small. Yet, as ³Whe validity, meaningfulness, and insights generated from qualitative
inquiry have more to do with the information-richness and the observational/analytical capabilities of
WKHUHVHDUFKHUWKDQZLWKWKHVDPSOHVL]H ́ (Patton, 1990, p.185), this small sample for expert validation
was considered acceptable. All the interview participants were senior executives that had on average
10 years of management experience. The sample for the content validation was completely male and
40 years and older. All had a post-graduate degree. A semi-structured interview method was used and
considered appropriate given that the limited existing information on strategic leadership did not allow
for rigidly structured interview schedules (Crabtree & Miller, 1999).
The content experts were asked to consider representativeness and clarity (Grant & Davis, 1997) when
scrutinizing the leadership measurement instrument. Representativeness meant the degree to which
each item was a reflection and operationalized the category to which it belonged (for instance whether