Ten quick tips for getting the most scientific value out of numerical data

Author summary Data expressed as numbers are ubiquitous in research in the life sciences and other fields. The typical scientific workflow using such numerical data consists of analyzing the raw data to obtain numerical results, followed by interpreting the results and presenting derived findings. In this article, we present some tips and tricks for working with numerical data. We discuss how data can be checked effectively for plausibility—and that this should be done frequently to spot potential mistakes early. Keeping clear records of the way raw numbers and intermediate or final results were obtained helps correct mistakes efficiently. To facilitate correct interpretation of data, we suggest paying attention to correct units of values and to understandable formulas. Appropriate statistical methods should be used for the specific question at hand. To avoid an over-interpretation of numbers, uncertainty of numbers should be respected, and one should be aware of cases in which bare numbers are actually misleading. Finally, we provide some recommendations for presenting data in a way useful for supporting interpretation and findings, ideally so that follow-up work can make use of the data.See it on Scoop.it, via Viruses, Immunology & Bioinformatics from Virology.uvic.ca
Ten quick tips for getting the most scientific value out of numerical data
Source: Viral Bioinformatics