Epidemiology & Biostatistics – Week 1 Lecture Notes

Introduction to the Unit

• Lecturer / Unit coordinator: Dr. Srikshanda
• Week 1 provides an orientation; students should also watch the separate orientation video.
• Aim: supply knowledge & practical skills to appraise and evaluate health-related evidence so that you can practise Evidence-Based Medicine / Evidence-Based Health Care (EBHC).
• Being an informed health professional is essential, not optional, in contemporary practice.

Why Evidence-Based Health Care Matters

• EBHC = management of patients/clients using the current best evidence of effectiveness.
• 5 classic EBHC steps
  • 1 – Identify an unanswered question/uncertainty.
  • 2 – Search current literature to see whether an answer already exists.
  • 3 – Critically appraise the located evidence (validity, reliability, applicability).
  • 4 – Apply the best evidence in practice.
  • 5 – Evaluate implementation outcomes later and feed findings back into practice.
• Key appraisal criteria
  • Validity – truth/accuracy of results.
  • Reliability – consistency/repeatability if the study were replicated.
  • Applicability – relevance to your local patient or client population.
• Three EBHC building blocks:
  1. Best available evidence.
  2. Clinical expertise.
  3. Patient values & circumstances.
• This unit specifically trains you to determine “what is the best available evidence?”

Unit Learning Outcomes

• Analyse different data types and apply appropriate statistical tests.
• Understand how information is generated through epidemiological study designs.
• Perform and interpret statistical inference – drawing conclusions about populations from samples.
• Differentiate between parametric vs non-parametric tests and know when each is suitable.
• Evaluate the accuracy of screening and diagnostic tests (sensitivity, specificity, predictive values, etc.).
• Ultimately:
  • Read and interpret published tables & graphs quickly.
  • Link statistical outputs to study conclusions.
  • Streamline literature appraisal for other units, future research projects, or clinical work.

Practical / Career Relevance

• Skills gained help with:
  • Tackling large volumes of literature for assignments.
  • Designing & executing research projects in later years (Honours, Capstone, etc.).
  • Future roles as clinicians, policy-makers, or (even) researchers.
• Statistics is a modern life skill – virtually every paper contains statistical reasoning.

Weekly Commitments & Study Advice

• Cramming does not work. Adopt “regular small doses”:
  • Listen to weekly lecture recordings (Epi + Biostats).
  • Attempt the weekly online quiz (details under Assessment).
  • Attend lab (computing) + tutorial (theory) OR, if fully online, work through equivalent material.
  • Join the open Collaborate sessions for Q&A; ideal for clarifying topics early (don’t wait until assessments).
• Table‐reading ability example: spotting significant differences using the $p$-value column (p<0.05 usually indicates a non-chance difference).
  • Example variables compared between heavy drinkers vs low-risk drinkers: experienced harm, second-hand harm, witnessed harm, academic problems, network of heavy-drinking friends.

This Week’s Specific Learning Objectives

• Define Epidemiology & Biostatistics and explain how they intertwine.
• Recognise the three EBHC elements already introduced.
• Master foundational terminology (population, sample, parameter, variable, etc.).
• Understand the four measurement scales (nominal, ordinal, interval, ratio) and why they dictate:
  • Choice of descriptive statistics.
  • Appropriate graphs.
  • Correct inferential tests.

Epidemiology in Depth

• Classic definition: “The study of the distribution and determinants of disease (or other health-related states/events) in specified populations, and the application of this study to the control of health problems.”
  • Remember the 3 D’s: Disease, Distribution, Determinants.
• Epidemiology vs Clinical Medicine:
  • Clinical – treats individuals (“Why did this patient get ill & how do we treat them?”).
  • Epidemiology – studies groups (“Which kinds of people get the disease & why?”; prevention focus).
• Core comparative study designs (each covered in upcoming weeks):
  • Case–Control – compare people with vs without disease (looks backward to exposures).
  • Cohort – follow exposed vs unexposed healthy groups over time to observe outcomes.
  • Experimental / Intervention / RCT – randomise to treatment vs control, then compare outcomes.
• Other epidemiological focus areas:
  • Measuring associations (correlation, causation).
  • Evaluating interventions, treatments, screening & diagnostic tests.
  • Describing disease patterns & trends, natural history, identifying high-risk groups, etc.

Biostatistics in Depth

• Biostatistics = collecting, summarising, analysing, and drawing objective conclusions from health-related data.
• Connection to Epidemiology:
  • Epi designs and gathers data.
  • Biostats turns that raw data into valid, reliable inferences.
• Two weekly lecture streams reflect this duality:
  • Epi lecture → concepts practised in tutorial exercises.
  • Biostats lecture → techniques practised in computing lab (SPSS).
• Together, they allow us to:
  • Exclude random chance as an explanation for findings.
  • Test theories & generate new knowledge.
  • Critically appraise published literature.

Real-World Applications

• Evaluate treatments, interventions, and health services.
• Validate screening/diagnostic tests (sensitivity, specificity, $\text{LR}^+$, $\text{LR}^-$, etc.).
• Compare disease & mortality rates across populations or time.
• Identify high-risk or high-need groups to allocate resources efficiently.
• Analyse temporal patterns (e.g., epidemic curves, seasonal trends).

The Generic Research Process (Road Map)

1. Start with a question / uncertainty.
2. Literature search to see if the answer already exists.
3. If not, plan & conduct a study:
   • Select suitable research design (influenced by question, resources, time).
   • Define target population & obtain a representative sample.
   • Decide what & how to measure (variables, instruments, timelines).
4. Data handling
   • Enter data into statistical software (SPSS for this unit).
   • Clean and screen (frequencies, graphs) for errors & outliers.
5. Choose appropriate statistical test(s) – must be suitable for the variable type and study design.
6. Analyse, interpret, and draw conclusions.
7. Disseminate (publish, present, implement).

Key Terminology — Foundations

• Population – entire group of people/objects you wish to study.
• Sample – subset chosen from the population for data collection.
• Sampling frame – list that enumerates every member of the target population.
• Simple random sample – every member has an equal probability of selection; requires a true sampling frame (common student misconception).
• Population parameter – unknown true value you wish to estimate (e.g., mean BP, prevalence).
• Descriptive statistics – summarise the sample (means, proportions).
• Inferential statistics – use the sample to infer about the population (hypothesis tests, confidence intervals).
• Variable – measurable characteristic that varies (height, pain score).
  • Independent variable / Exposure – presumed cause or determinant.
  • Dependent variable / Outcome – presumed effect.
  • Extraneous variable – controllable factor that could influence results.
  • Confounding variable – uncontrolled factor linked to both exposure and outcome.
  • Demographic variables – age, sex, education, etc.
• Sampling variation – natural fluctuation of estimates across different samples.
• Sampling error – difference between a sample estimate and the true population value, unavoidable because we rarely study whole populations.

Data Types & Measurement Scales

• Two overarching data families
  1. Categorical (Qualitative)
  2. Continuous (Quantitative)
• SPSS groups interval + ratio together under “Scale”.

1 – Categorical Data

• Individuals are placed into categories.
• Sub-classes:
  • Nominal (names only, no intrinsic order)
  • If only two categories ⇒ Binary Nominal (e.g., yes/no, male/female).
  • Ordinal (ordered / ranked categories, but unequal intervals)
  • E.g., disease severity: mild → moderate → severe.

2 – Continuous Data

• Numeric, measurable, inherent order, theoretically infinite precision.
• Sub-classes:
  • Interval – equal units; no true zero (e.g., temperature $^\circ\text{C}$, IQ).
  • Ratio – equal units with an absolute zero (e.g., weight, height, Kelvin, exam mark if no negative scores).

Why Scales Matter

• They determine:
  • Suitable descriptive measures (mean/SD vs median/IQR vs counts/percent).
  • Appropriate graphs (histogram, box-plot, bar chart, pie chart).
  • Valid inferential tests (e.g., $\chi^2$ for nominal vs $t$-test/ANOVA for normal-scale data).
• Incorrect test choice ⇒ no marks in assessments (“No output, no mark” rule).

Resources & Support

• Weekly Collaborate drop-in sessions commence Week 2.
• Discussion boards open for asynchronous questions.
• Core textbooks / articles (full list in Blackboard).
• SPSS installed on campus machines; instructions supplied in lab sheets.

Final Reminders

• Keep up with weekly content—small, regular practice beats cramming.
• Watch orientation + week 1 videos, then attempt Quiz 1.
• Bring questions to tutorials, labs, and Collaborate; clarity early will save assessment stress later.
• Best wishes for a productive semester!