“…Moreover, the two conditions (based on firm and TMT size) that simultaneously had to be met are difficult to detect a priori. The snowball sampling procedure helps locate members of special hard‐to‐find populations via referral by network contacts (Biernacki and Waldorf, ; Saunders et al, ) and has already been used in family business studies that faced comparable constraints (e.g., Bettinelli, ; Björnberg and Nicholson, ; Farrington et al, ; Fiegener et al, ; Van der Merwe, ; Venter et al, ). The risk of sampling bias through this specific method (Lee, ), is less of a problem in this study because the descriptive statistics (more details in Table on p. 15) provide evidence of sufficient variation within the sample in terms of firms and respondents (e.g., firm size, firm age, top management composition).…”