In the first data collection wave, the Researcher carried out online surveys among three distinct groups of experienced managers and leaders. The chosen aproach followed broadly the principles of the so-called Critical Incident Technique (CIT), which has been first described by John C. Flanagan (1954). Click here to find the pdf version of the full article describing the methodology.

The free-text reponses obtained through the surveys were analysed qualitatively and quantitatively. The content analyses in five iterations resulted in seven high-level behavioural categories (see here). The appropriateness of these high-level categories was confirmed with the help of two groups of experienced professionals in an inter-rater agreement study with, in total, 104 individual statistical tests. Click here for a diagram that illustrates the approach.

Based on the validated behavioural categories, a pool of 168 items were written from scratch, each item being linked to one or two concrete behaviours. This item pool was validated in a further inter-rater agreement study, and the number of items was reduced by approximately 50% to 80. This set of items was then subjected to an empirical analysis based on a pilot study and a subsequent larger-scale development sample. Item- and factor analytical procedures were applied and the emerging factor solution successfully tested for "simple structure" (Thurstone, 1947). Several plausible, yet competing, structural models, were compared with each other using confirmatory factor analysis techniques.