Types of Data & Methods of Representation

Types of Data & Methods of Representation

1. Definition & why data type drives every downstream choice

• Data = the recorded values of one or more variables measured on a set of subjects/observations; the very first step before any calculation or plotting is to decide what type of data one is dealing with, because the type dictates the appropriate summary statistic, the correct graph, and the valid statistical test (Swinscow 10e Ch.1, p.1).
• A variable is any characteristic that can take different values between subjects (e.g. height, blood pressure, sex, breast-cancer grade); the question "what type?" must be asked of every variable before analysis (Swinscow 10e Ch.1, p.1).
• Biostatistics = statistics applied to the health and medical sciences — the science of gathering, summarising, and analysing biological/medical data; it differs from core statistics by being driven by a study hypothesis → data gathering → analysis → inference cycle, so correct data classification is foundational to designing the experiment and choosing its analysis (Medhi Ch.11, p.123).
• Population vs sample — a population is the entire set of subjects of interest (e.g. all AIIMS/PGIMER patients); a sample is a randomly selected subset that represents (but never replicates) the population; the gap between a sample value and the true population value is the sampling error (Medhi Ch.11, p.123).
• Why classification matters clinically: the data type determines whether a mean ± SD or a median (IQR) or a proportion is the right summary, whether a parametric or non-parametric test applies, and whether a histogram, bar diagram, box-plot, or scatter plot is the honest visual (Swinscow 10e Ch.1, pp.1–2; Medhi Ch.11, p.125).

Provenance / source-of-data sub-classification (Medhi Ch.11, p.124):

• Primary data — collected first-hand directly from the research experiment.
• Secondary data — obtained from other/indirect sources (published literature, registries, another investigator's dataset).