دانلود رایگان مقاله انگلیسی یک راهنمای عملی برای کلان داده - الزویر 2018

Abstract

Big Data is increasingly prevalent in science and data analysis. We provide a short tutorial for adapting to these changes and making the necessary adjustments to the academic culture to keep Biostatistics truly impactful in scientific research.

7 The incentive system

Alas, all is for naught if the incentive system is outdated. Indeed, junior investigators are smart and understand exactly what is expected for fast and painless promotion. Most of them are told that they need to publish many papers in the top 5 or 6 Bio/Statistical journals, which favor theory. Thus, the most creative 10 to 12 year period of the life of a Bio/Statistician is spent on things that have nothing to do with data, in general, or Big Data in particular. Indeed, some Departments go as far as placing the Annals of Statistics above other Statistical journals, most likely because it is a more theoretical journal. This message is unmistakable and can only lead to more junior investigators being disengaged from data, be it big, small or, even, moderate.

Instead, the incentive system needs to evolve, become more flexible, and more inclusive. In particular, Departments should encourage their faculty to publish in top tier journals, irrespective of the area of research. The idea that publishing in applied journals is easier should be abandoned and replaced by the requirement for the Bio/Statistician to be a leader in whatever area of scientific research they choose to work in. For example, if a Biostatistican works on nutrition research then they should participate in the most important conferences in nutrition research, be recognized in, contribute to, and drive the methodological research underlying nutrition research. In this context Statistical novelty may not be an absolutely new model or theory, but a carefuly crafted, targeted, highly impactful contribution to a new area of science. Novelty does not spring only from completely new methods, but also from carefully tuning and polishing existent methods for entire new areas of science and disseminating these approaches to our collaborators and colleagues.