Biostatistics: Basic Concepts and Applications

• Outlines of biostatistics include origin, development, aims, history, definitions, reasons to know, philosophy, theory of knowledge, research methodology, causality, estimation, population, sampling approaches, and data presentation methods.
• Course content will focus on practical statistical knowledge and applications rather than theorems and proofs.

Books

• Basic & Clinical Biostatistics by Beth Dawson, Robert G. Trapp is a biostatistics textbook
• Clinical Epidemiology & Biostatistics by Rebecca G. Knapp, M. Clinton Miller III is another biostatistics textbook
• Temel ve Klinik Biyoistatistik by Prof.Dr. Rian Dişçi is a biostatistics textbook

Lecture Programs Include

• Basic concepts in clinical research, uncertainty, significance, confidence intervals, errors (Type-I, Type-II, BIAS), theory of knowledge, terminology, and variable types and measurements
• Sampling and sample size calculation.
• Data summarization with central tendencies and distribution measures like mean, median, mode, SD, SE, and CV
• Probability and normal distribution.
• Hypothesis testing using parametric and nonparametric methods (t-test, ANOVA, Mann-Whitney)
• Categorical data Analysis (Chi-square, Fisher, etc.) and Qualitative relationship/symmetry/agreement evaluation methods (Mc-Nemar, Kappa, etc.)
• Correlation and Regression Analysis
• Diagnostic Tests (Decision Making) and ROC curve
• Risk Analysis including RR (Relative Risk), OR (Odds Ratio), Epidemiological measures, Ratio, proportion, rate, prevalence, incidence, etc
• Survival Analysis (Life Table, Kaplan-Meier and Log-rank test)
• Clinical trials; History, Research Designs, Protocol, Classifications and Strengths of Evidence,Good Clinical Practice and Guide Lines; Equator Network; CONSORT, ARRIVE, etc.

Introduction to Research

• A systematized effort to gain new knowledge
• Research = Re+Search (Seeking facts or principles)
• A search for knowledge through objective methods of finding solutions to a problem or of systematic approach concerning generalization and formulation of a theory
• Research is the process of identifying and solving problems

Why Conduct Research

• To discover answers to questions through applications of scientific procedures
• To find out hidden truths that have not been discovered yet

Introduction

• The aim of researchers is to demonstrate the validity of an observation and conduct hypothesis tests that help make decisions about population characteristics.
• Limited sample information helps obtain accurate inferences about the population.
• Definitions and statistical estimates generally have a confidence level of 95%.

Scientific Method

• The way researchers use knowledge and evidence to reach objective conclusions about the real world
• The analysis and interpretation of empirical evidence to confirm or disprove prior conceptions

Characteristics

• Public, objective, empirical, systematic, cumulative, includes theory, law, predictive and self-correcting
• Scientific Advances rely upon freely available information
• Science tries to rule out judgment eccentricities by researchers and institutions
• Researchers focus on a knowable and measurable world
• Metaphysical, nonsensical explanations of events are rejected
• Research follows a specific method

Elements

• A set of related propositions that systematically views phenomena by specifying relationships among concepts is part of theory
• A statement of fact concisely explains an action generally accepted as true and universal is a law
• Science relates the present to the future by making predictions
• Corrections of thoughts, theories, or laws are appropriate when errors in previous research are uncovered.

The scientific method

• Objectively establishing facts through testing and experimentation involving making an observation, forming a hypothesis, making a prediction, conducting an experiment and analyzing the results

Stages

• Experience and knowledge are necessary to raise a research question and to be aware of causality
• Description of the facts and problems involves metaphysical and concrete systematization of the related processes of the research topic and causality
• Planning the research design involves thinking and systematizing the operations necessary to achieve a desired goal, creating a hypothesis (Null/No Difference Hypothesis, Proposed/Difference Hypothesis), sampling, and sample size calculation
• Determination of primary variables, data collection (sampling and measurement methods) are key
• Verification involves testing the hypothesis, selecting the right biostatistical methods, and data processing/analysis
• Generalization concludes on hypothesis checking and reporting, sharing research findings with the field of science

The Origin and Development of Statistics in Medical Research

• In 1929, a significant paper on the application of statistics was published in the Physiology Journal by Dunn
• By 1937, Austin Bradford Hill had published 15 articles on statistical methods in book form
• In 1948, a randomized controlled trial (RCT) of Streptomycin for pulmonary tuberculosis was published, in which Bradford Hill played a key role
• From 1952 to 1982, there was an 8-fold increase in the growth of Statistics in Medicine

Statistics

• The science of collecting, monitoring, analyzing, summarizing, and interpreting data

Biostatistics

• Applied to biological problems, including public health, medicine, and ecological and environmental contexts
• More than just biology, biostatisticians must also learn biology

By Definition

• Biostatistics studies the collection, organization, and summarization of biological data to draw inferences, evaluate conclusions, and present results.
• Practical activity that involves collecting, organizing, processing and analyzing statistics
• Scientific methodology with modeling purposes
• Totality of the data collected and systematized
• This is multitude where elements of the population are the same according to given characteristics, while based on other parameters they are different.
• Involves collecting, processing and analyzing information on individuals of mass phenomena, It is a concise quantitative characterization on the analyzed phenomena

The Role of Biostatisticians

• Guide the design of experiments or surveys before data collection, including sample size, power calculations, sample selection, control groups, and designing questionnaires
• Analyze data using proper statistical procedures and techniques (data management, descriptive statistics, graphs, and univariate and multivariate statistical analysis)
• Present and interpret results to researchers and decision-makers

Why is Biostatistics Important?

• It allows extrapolation from collected data to make general conclusions about a larger population
• General conclusions are possible from small data samples
• This process works by assuming all data is sampled randomly and infinitely
• It then analyzes this sample to make inferences about the larger population.

Statistics: The Art and Science of Data

• Deals with planning research, collecting data, summarizing data, analyzing data, interpreting findings (reaching decisions or discovering new knowledge), and presenting results.
• Much of life has a systematic and random component
• Smoking and lung cancer are associated but not causal
• There are systematic components associated with these things

Analysis

• Starts with defining a population of interest
• Randomly selects sample of subjects to study
• Subdivides the subjects; One half recieves the treament and the other hlaf something else
• Measure variables from baseline for both groups and trail outcome variables within both group
• Statistical techniques make interences about both distributions in the general population in addition to the effect of the treatment

Challenge

• A lot of life is made up of a systematic component and a random component
• Example:
  • Smoking is associated with lung cancer.
  • Everyone that smokes doesn't get lung cancer, and not everyone that gets lung cancer smokes
  • There is still an association though
• Sample size is needed demonstrate which group is more significant
• Need to evalute treatment as better and the level
• Measurement variation including mild, moderate and severe need to be explored
• Reliablity and validity need to be confirmed
• Magnitude lab errors need to be found and what is acceptable
• How to interpret abnormal values need to be decided