When entrepreneurial performance is plotted as a distribution, it is often right-skewed and heavy-tailed. In other words, a few star performers typically dominate most of their peers in entrepreneurial contexts. Thus, a simplistic focus on average entrepreneurial performance – while ignoring star performance – can limit our empirical and theoretical understanding.
The availability of distributional information in entrepreneurship research
There remains limited reporting of distributional information (e.g., skewness, kurtosis, and shapes) for performance in empirical studies published in high-impact entrepreneurship journals. However, commonly reported descriptive statistics (i.e., sample size, mean, and standard deviation) can be used to reveal insights into the prevalence of star performance. For example, the coefficient of variation (CoV), which is simply the ratio of the standard deviation to the mean, can be computed to infer the extent of volatility in performance. Moreover, if skewness of performance is reported, it can be used in conjunction with the CoV to create a moment-ratio plot and thus identify the most likely distributional shape (e.g., lognormal) for entrepreneurial performance. Such heuristics can serve as a starting point for explicitly examining the entire range of performance.
Reverse engineering distributional information using simulations
Because most studies of entrepreneurial performance neither report skewness nor share the raw data, we use a simulation-based methodology in our paper published in the Journal of Business Venturing Insights to obtain rough estimates of skewness. Here, we first test the simulation methodology using statistics reported in two studies that explicitly examine performance distributions. Next, we apply the methodology to 122 measures of performance reported in 93 papers published in three high-impact entrepreneurship journals. We find that the vast majority of performance outcomes have skewness greater than 2, indicating the prevalence of star performers.
What can researchers do?
Our central argument is that entrepreneurship research should explicitly examine the entire range of performance by adopting a distributional perspective. Empirical studies should report not only the median, minimum, and maximum of each performance measure but also its skewness, kurtosis, and likely distributional shape. Moreover, we show how heuristics and simulations can enrich our shared understanding of star performance when raw performance data is not reported or is subject to confidentiality.
What can practitioners do?
Entrepreneurship practitioners – whether educators, executives, founders, investors, or policy-makers – can enrich their decision-making by shifting their focus from average outcomes to the entire range of performance, with an emphasis on influential outliers. By thus adopting a distributional perspective, practitioners can draw upon emerging entrepreneurship research to better understand and predict star performance.
Read the full paper here to find out more: https://www.sciencedirect.com/science/article/pii/S2352673424000441
Author bio
Kaushik Gala is a PhD Candidate in the Department of Management and Entrepreneurship, Ivy College of Business, Iowa State University. His research interests include star performers, distributions of entrepreneurial performance, and digital platform entrepreneurship. He is also interested in entrepreneurial cognition and corporate venture capital.
Andreas Schwab is a Professor in the Department of Management and Entrepreneurship, Ivy College of Business, Iowa State University. His research expertise includes corporate entrepreneurship, innovation and organizational learning, entrepreneurship in digital-platform ecosystems, and research methodology.
Leave a Reply