Comprehensive Introduction to Biostatistics and Descriptive Methods

Foundations and Definitions of Biostatistics

  • Definitions of Statistics:

    • Mathematics of experimentation (K. Mather).

    • Scientific study of numerical data describing natural variations (Sokal & Rohlf, 1981).

    • Set of methods used to collect, organize, summarize, present, and analyze data to draw conclusions and make decisions.

    • Mathematical and scientific instruments used to determine characteristics of a data set.

  • Biostatistics: Application of statistical tools to biological, medical, and public health data to guide decisions under uncertainty.

  • Domains of Application: Medicine, epidemiology, demography, geophysics, economics, social sciences, marketing, and physics.

  • Core Branches:

    • Descriptive Statistics: Techniques to organize, summarize, and display data via numerical indices and visual tables/graphs.

    • Inferential Statistics: Drawing conclusions about an entire population from a representative sample using hypothesis testing (null vs. alternative), confidence intervals, and p-valuesp\text{-values}; common tests include the t-testt\text{-test}, chi-square test, ANOVA, and regression.

Stages of Statistical Investigation and Core Terminology

  • Five Investigation Stages:

    1. Data Collection: Defining characteristics, population, sampling method, and study protocols.

    2. Data Organization: Editing, classifying, and tabulating gathered data.

    3. Data Presentation: Visualizing patterns via tables and charts.

    4. Data Analysis: Computing statistical indicators such as the mean, median, mode, range, and variance.

    5. Data Interpretation: Drawing meaningful conclusions from analytical results.

  • Fundamental Terminology:

    • Population: Total collection of units (individuals, records, events) sharing at least one common characteristic.

    • Sample: Representative subset of a population selected by chance to generalize findings.

    • Unit of Observation: Elementary unit of a population evaluated for a specific characteristic.

    • Observation: Action of identifying or measuring a variable on an observational unit (exhaustive or partial).

    • Rate: Probability of an event occurring over time:

Rate=Number of events occurring during [t1,t2]Population exposed to the risk of the event\text{Rate} = \frac{\text{Number of events occurring during } [t_1, t_2]}{\text{Population exposed to the risk of the event}}

Biological and Metrological Variability

  • Biology as an Observational Science: Biological phenomena are subject to natural variation rather than exact determinism.

  • Components of Global Variability:

    • Biological Variability:

    • Intra-individual variability: Variations occurring within the same individual over time.

    • Inter-individual variability: Differences observed between different individuals.

    • Metrological / Experimental Variability: Inconsistencies introduced by measurement instruments, laboratory devices, or experimental execution.

Classification of Variables

  • Variable: Characteristic or attribute that assumes different values across observational units.

  • Modality: Specific value or category that a variable can take.

  • Qualitative (Categorical) Variables:

    • Nominal: Categories that cannot be logically ordered or ranked (e.g., eye color, religion, sex, place of residence).

    • Ordinal: Categories that possess an inherent order or hierarchy (e.g., socioeconomic status, disease severity, blood pressure classification).

  • Quantitative (Numerical) Variables:

    • Discrete: Expressed exclusively as whole numbers or counts (e.g., number of children, cell counts).

    • Continuous: Takes any real value within a continuous interval (e.g., age, height, blood pressure levels).

Frequency Distributions and Continuous Data Grouping

  • Frequencies:

    • Absolute Frequency: Direct count of units in a given category.

    • Relative Frequency:

\text{Relative Frequency} = \frac{\text{Absolute frequency of the group}}{\text{Total population frequency}}}

  • Cumulative Headcounts: Classified into "less than" (accumulating from the first/lowest category upward) and "more than" (accumulating from the highest category downward).

    • Class Intervals for Continuous Data:

  • Must be successive, contiguous, mutually exclusive, non-overlapping, and collectively exhaustive.

  • Boundary Rule: The lower boundary is included; the upper boundary is excluded.

  • Class Width:

Width=Upper Boundary−Lower Boundary\text{Width} = \text{Upper Boundary} - \text{Lower Boundary}

  • Class Midpoint:

Class Midpoint=Lower Boundary+Upper Boundary2\text{Class Midpoint} = \frac{\text{Lower Boundary} + \text{Upper Boundary}}{2}

  • Range (Span / Domain of Variation):

Range=Maximum Value−Minimum Value\text{Range} = \text{Maximum Value} - \text{Minimum Value}

  • Number of Classes (CC): Standardly set between 55 and 2020 using:

C=NC = \sqrt{N}

C=1+3.3log⁡10(N)C = 1 + 3.3 \log_{10}(N)

Tabular and Graphical Data Presentations

  • Tabular Representations:

    • Single-Entry (Unidimensional) Table: Details the distribution of a single variable.

    • Double-Entry (Two-Dimensional) Table: Displays joint distributions, such as contingency tables for two qualitative variables or comparison tables for one qualitative and one quantitative variable.

  • Graphical Representations:

    • Qualitative Variables:

    • Bar Chart: Disconnected bars with constant arbitrary width and heights/areas proportional to frequencies.

    • Pie Chart: Slices proportional to category frequencies (100%=360∘100\% = 360^\circ).

    • Discrete Quantitative Variables: Discontinuous bar charts.

    • Continuous Quantitative Variables: Histogram featuring contiguous, juxtaposed rectangular areas without gaps.

Questions & Discussion

  • Question: Which of the following statements defines "Statistics"?

    • Answer: The mathematics of experimentation; the scientific study of numerical data describing natural variations; and a set of methods used to collect, organize, summarize, present, and analyze data to draw conclusions.

  • Question: What are the main application areas of statistics in health and society?

    • Answer: Medicine and epidemiology, demography, and geophysics.

  • Question: What are the two primary branches of statistics?

    • Answer: Descriptive statistics and inferential statistics.

  • Question: The nominal variable is a:

    • Answer: Qualitative variable that cannot be ordered.

  • Question: Which of the following is an example of a nominal variable?

    • Answer: Whether or not the medicine is effective.

  • Question: The ordinal variable is a:

    • Answer: Qualitative variable that can be ordered.

  • Exercise: Analysis of 159159 incidental intracranial aneurysms (IIAs\text{IIAs}):

    • Variable & Type: Size of IIAs\text{IIAs} (in millimeters); quantitative continuous variable.

    • Number of patients with size between 10−14 mm10 - 14\,\text{mm}:

159×0.163=26159 \times 0.163 = 26

  • Number of patients with size ≤19 mm\le 19\,\text{mm}:

29+87+26+10=15229 + 87 + 26 + 10 = 152

  • Relative frequency of size between 5−9 mm5 - 9\,\text{mm}:

87159≈0.547\frac{87}{159} \approx 0.547

  • Proportion of patients with size <15 mm< 15\,\text{mm}:

29+87+26159=142159≈0.893\frac{29 + 87 + 26}{159} = \frac{142}{159} \approx 0.893

  • Percentage of patients with size between 15−29 mm15 - 29\,\text{mm}:

10+4+1159×100%=15159×100%≈9.43%\frac{10 + 4 + 1}{159} \times 100\% = \frac{15}{159} \times 100\% \approx 9.43\%

  • True class interval of (20−24)(20 - 24): [19.5,24.5)[19.5, 24.5) (or [20,25)[20, 25)).

  • Class Width: 5 mm5\,\text{mm}.

  • Maximum value: Located in class interval 30−34 mm30 - 34\,\text{mm} (34 mm34\,\text{mm} limit).

Is this all ?is this all