Desktop version

Home arrow Business & Finance

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Survey Design

For the purpose of this investigation, the survey design for organizational leaders consisted of socio-demographic questions. Assessing physical distance, one question was included to retrieve leaders’ permanent work location.

The survey for followers consisted of six sections, covering self-ratings of selfleadership behavior, other-assessment of perceptions of FRL dimensions of supervisors, LMX, interaction frequency, performance assessment, and sociodemographic data. Additionally, permanent office location was assessed. With a total number of 84 items, the completion time was estimated at 20 minutes. Sections of the survey are outlined in Table 9. Excerpts of the questionnaire are illustrated in Appendix A.

Table 9. Composition of the Survey Instrument




Number of items


Leading myself


Furtner & Rauthmann (in prep.)



My leader and me

MLQ 5X short

Bass & Avolio (1995)



Myself and my leader


Graen & Uhl-Bien (1995)



Interaction frequency

Andressen et al. (2012)




Walumbwa et al. (2008); Heilman, Block & Lucas (1992)



Descriptive information


Total number of items


Content validity and preliminary analysis

Prior to data collection, the content of the research instrument was examined in a first step to test and, if necessary, enhance validity. Augmenting validity aims at ensuring that measures do assess what they are supposed to (Balnaves & Caputi, 2001). As various survey instruments with multi-dimensional factors were deployed, a method developed by Lawshe (1975) was utilized. Content validity tests for relevance of each item to represent a certain construct (Haynes, Richard & Ku- bany, 1995, p. 239). The researchers define content validation as “a multimethod, quantitative and qualitative process that is applicable to all elements of an assessment instrument” (Haynes et al., 1995, p. 247). The purpose of content validity is to narrow down potential errors associated with the research instrument. Weiber and Muhlhaus (2014, p. 157) consider content validity to exist when indicators capture the entire semantic meaning of a construct. The analysis was conducted through a panel of experts who were asked to rate each construct item in terms of necessity, assessing the relation between variables. If more than 50% of the judges rate the items as essential, content validity is confirmed. Particular criticism has been expressed that the MLQ 5X suffered from low content validity (Schriesheim, Wu & Scandura, 2009). Findings propose careful application of the MLQ on various levels. The authors recommend the revision of the MLQ to make it a valid instrument applicable to more levels.

The panel consisted of five research experts, organizational leaders, and market research experts who were otherwise not involved in the study. The specialists were provided with hard copies of the MLQ 5X short, SLSI, LMX-7, performance and interaction frequency scale, as well as all demographic measures.

Individuals who tested the instrument for content validity stated slight skepticism with items of the SLSI. Repetitive, similarly sounding items were increasingly frustrating to the participants. One participant stated “I am confused and a little frustrated. It sounds to me like the essence of the questionnaire has been on interpretation of wording in the questions than the real behavioral aspects of me”. Yet, most of the respondents regarded the items as relevant and recommended to keep them.

Following the tests on content validity, a preliminary test was conducted to analyze the research instrument for possible limitations in design (Cooper & Schindler, 2008). Prior to distribution to participants, the survey instrument was handed out to 25 students enrolled in a Master’s program. In addition to understanding the formulation of the items, the survey instrument was tested and Cronbach alpha coefficients were computed. Reliability scores for MLQ 5X short, SLSI, LMX-7, and performance scale were lying in a range between .71 and .94 except for the first- order factor of laissez-faire leadership which resulted in a fairly low value of .54. Consequently, the four items of the laissez faire scale were recoded to ensure a straightforward positivistic embodiment of wording.

<<   CONTENTS   >>

Related topics