Statistical considerations for research

Abstract

The ultimate objective of research on injuries is to generate knowledge that could be used to reduce trauma and injuries worldwide. Knowledge is generated from interpreting and contextualizing information, while information, both qualitative and quantitative, comes from synthesizing data. The science of statistics provides researchers with the necessary tools to quantify the variability in the observations, as well as identify patterns and relationships in quantitative data. Many folks think of statistical sciences as a branch of mathematical sciences. The science of statistics relies on mathematical sciences to handle quantitative data, but it is closer to philosophy in that it aims to understand nature and its processes, relationships, characteristics and patterns (see Figure 1). It postulates or hypothesizes these characteristics, abstractly, theoretically, but then desires to verify if these postulates are true. Statistical science accepts that the truth can never be known with certainty. Thus, statistical thinking differs conceptually from mathematical thinking by recognizing that there is inherent variability in nature, and that chance plays an important role in what is observed. Mathematics is a necessary tool for statistics, just like a hammer and a saw are necessary tools for a carpenter. However, the application of statistical thinking and methods helps researchers understand nature and its processes amidst uncertainty. Statistics focuses on quantifications or measurements. Thus, four statistical quantification aspects must be considered - quantifying uncertainty or variability, quantifying probability or chance, quantifying risk and exposure, and quantifying the strength of relationships – all of which are essential elements of injury prevention and control research (Bangdiwala & Banerjee Taylor, 2011).