The Editor-in-Chief of a scientific journal is wondering how to reduce the review cycle times to offer timely feedback to authors and improve the stature of the journal. The journal is fictional, but the data is real. Through the analysis, it appears that with parallel activities having random durations, one should not trust averages because one is constrained by the slowest activity. Sources of delay are explored, such as Parkinson’s Law, the Student Syndrome, and the Planning Fallacy.
Understand the sources and performance impact of variability in knowledge-intensive processes. Specifically, o Variability is often man-made, i.e., there are no physics governing process variability (unlike manufacturing). Moreover, it can be substantial. o With parallel processing, averages should not be trusted. o Parkinson’s law: Work expands to fill the time available. o Student Syndrome: Start as late as possible. o Planning Fallacy o With variability, characterizing performance should be done in probabilistic terms, i.e., characterizing performance only in terms of its mean is often too restrictive. o Need to develop an exception handling process and be transparent about it. Delve into (or revisit) core statistics concepts: o Correlation in sequential activities affects variability, not mean processing times. o If variance of the sum of random variables is less than the sum of their variances, correlation must be negative. o Jensen’s inequality: The expectation of a convex function of random variables is greater than the function of their expectations. Discuss the incentives and behavioral mechanisms at work when managing time in knowledge-intensive processes when monetary payments (i.e., financial penalties for being late) are not possible. o Realize that only process owners truly care about operational performance; other stakeholders are willing to contribute out of feelings of reciprocity.
- Process analysis
- project management
- variability
- incentives
- behavioral biases
- Q12025
