Construct measurement in strategic management research: illusion or reality?

Dance with the one that brought you? Venture capital firms and the retention of founder‐CEOs

“…Further, combining these three indicators into single index avoided collinearity problems associated with the high correlation between prior CEO and prior start-up experience. Our approach was also consistent with recent calls for greater use of indices in strategy and governance research to operationalize multidimensional constructs (Boyd, Gove, and Hitt, 2005). Number of prior Internet-industry deals.…”

A socioemotional wealth perspective on innovativeness and performance of family businesses.

“…The used input parameters were: effect size (r = .20), desired power level (.80) and alpha (.05). The value of 0.80 has been the recommended threshold of power adequacy (Cohen, 1988) and the one that has been mostly used in prior research (Boyd, Gove, and Hitt, 2005). The power analysis results indicated that a sample size of 146 participants would be required for the purposes of the main analysis in order to achieve adequate power of 0.80 or, in other words, in order to have an 8 in 10 chance to successfully detect a relationship that exists.…”

Causation, counterfactuals, and competitive advantage

“…1 Second, positivist research relies on statistical methods and tests. Nevertheless, observational biases, measurement errors, model misspecification, and the ambiguity of findings compromise the ascertainment of causation (Bergh and Holbein, 1997;Boyd et al, 2005b;Denrell, 2003;Shaver, 1998). For instance, a direct empirical association of resources with superior financial performance does not preclude omission or misspecification of links in the complex causal chain, such as industrial conditions, resource properties, competitive advantage, or superior performance (Cockburn et al, 2000; see also Tsang, 2006 about assumption-omitted testing).…”