MATH2209 Biostatistics
The Complete Study & Assessment Guide · S2 2026MATH2209 Overview
- RMIT University
- Semester 2 2026
- Level 2 statistics
- Eleven teaching modules
- Excel, SPSS and R
What MATH2209 teaches
Biostatistics applies statistical reasoning to biological, medical and epidemiological questions. The course begins by classifying variables and describing data, then uses probability to model uncertainty.
- Assessed through Module assignments, two on-campus summative tests and a staged statistical project.
- Most important habit Choose the method from variable type and study design before calculating.
- Where marks depend Show equations, substitutions and interpretation rather than an unsupported final answer.
- How to revise Mix competing methods so each question forces a fresh selection decision.
What you get in the full MATH2209 Study & Assessment Guide
- 11 chapters, 47 pages — every assessed topic, week by week
- 31 original practice questions, each with a model answer
- 11 worked examples, solved the way MATH2209 marks
- 22 computed figures, redrawn from the course’s own numbers
- Before-you-submit checklist for the Final (25%) — what markers reward, what costs marks
Free on this page: assessment table, chapter map, one worked example, key terms, FAQ. The rest is in the full study & assessment guide.
How MATH2209 is assessed
| Component | Weight | Format |
|---|---|---|
| Assessment task 1 - Module Assignments | 20% | Untimed module activities and practice questions; non-multiple-choice answers require full working |
| Mid-semester Summative Assessment 1 | 15% | On campus, Modules 1 to 4, 90 minutes, one attempt, online with permitted notes and calculator |
| Final: Summative Assessment 2 | 25% | On campus and cumulative across Modules 1 to 11; confirm current operational details on Canvas |
| Project Proposal | 5% | Statistical investigation proposal using the provided template |
| Project Data File | 5% | Finalised analysis-ready data file with a clear dataset description |
| Project Poster | 25% | Scientific poster summarising the investigation, analysis and findings |
| Project Video Presentation | 5% | Three-minute informative presentation of project results |
Weights come from the Semester 2 2026 Canvas Assessment Summary and total 100%. The syllabus labels the 25% final component Final Test: Assessment 2, while the assessment summary labels it Final: Summative Assessment 2. Confirm current dates, location, permitted materials and submission instructions on Canvas.
Assessment structure
Bar widths reproduce the seven published weights and sum to 100%.
What other students are asking about MATH2209
- Where can I find lecture notes or study guides for MATH2209 at rmit?
- The MATH2209 study guide in AskSia Library covers Biostatistics in 11 chapters, with the assessment table, a chapter-by-chapter topic map, key terms and a worked example, free to read on this page.
- Is there a final exam in MATH2209?
- No final exam is listed in the offering this AskSia Library guide is built from; MATH2209 is assessed by Assessment task 1 - Module Assignments 20%, Mid-semester Summative Assessment 1 15%, Final: Summative Assessment 2 25%, Project Proposal 5%, Project Data File 5%, Project Poster 25%, Project Video Presentation 5%.
- What does MATH2209 cover on probability rules and bayes?
- In AskSia Library's MATH2209 guide, the Probability Rules and Bayes chapter explains complements, unions, intersections, conditional probability, independence and Bayes' rule.
- What does MATH2209 cover on binomial and poisson models?
- In AskSia Library's MATH2209 guide, the Binomial and Poisson Models chapter explains fixed-trial success counts, event rates, probability calculations, means and standard deviations.
- What does MATH2209 cover on normal distributions and z scores?
- In AskSia Library's MATH2209 guide, the Normal Distributions and Z Scores chapter explains normal curves, standardisation, tail areas, percentiles and inverse calculations.
Current MATH2209 dates
| Date | Item | Control |
|---|---|---|
| 5 November 2026 | Final Test: Assessment 2 | The Semester 2 syllabus prints 9:30am to 11:30am and marks the details as subject to change; confirm Canvas. |
Dates are as published in the Semester 2 2026 Canvas syllabus. Confirm exact deadlines and submission settings in the live LMS.
What MATH2209 covers
The course moves from descriptive statistics and probability through distributions, estimation and hypothesis testing, then applies inference to independent, paired, ranked, categorical, regression and epidemiological settings. The chapter order follows the eleven published Canvas modules so each method arrives after the ideas it depends on.
Data, Variables and Descriptive Statistics
Variable types, measurement levels, centre, spread, distribution shape and epidemiological contextThis chapter begins with the decision that controls every descriptive calculation: what kind of variable was measured.
- Categorical and quantitative variables
- Nominal and ordinal measurement
- Discrete and continuous data
- Mean, median and mode
- Variance and standard deviation
- Distribution shape and outliers
Probability Rules and Bayes
Complements, unions, intersections, conditional probability, independence and Bayes' ruleProbability becomes easier when ordinary language is converted into events and reference groups. This chapter separates intersections, unions, complements, independence and mutual exclusivity before building conditional probability and Bayes' rule.
- Complements and unions
- Intersections and overlap
- Conditional probability
- Independent events
- Mutually exclusive events
- Bayes' rule and reversed conditions
Binomial and Poisson Models
Fixed-trial success counts, event rates, probability calculations, means and standard deviationsBoth distributions describe counts, but their data-generating mechanisms are different. The binomial model counts successes across a fixed number of independent binary trials with constant success probability.
- Binomial trial conditions
- Exact and cumulative counts
- Binomial mean and spread
- Poisson rates and exposure
- Poisson probability calculations
- Boundary language and tail checks
Normal Distributions and Z Scores
Normal curves, standardisation, tail areas, percentiles and inverse calculationsThe normal model describes a symmetric continuous distribution through its mean and standard deviation.
- Normal curve properties
- Standard normal scale
- Raw values and z scores
- Left and right tail areas
- Between-value probabilities
- Inverse normal calculations
Sampling, Estimation and Confidence Intervals
Sampling methods, bias, the central limit theorem, standard error and interval interpretationEstimation connects a sample statistic to an unknown population parameter while acknowledging sampling variability.
- Probability sampling methods
- Selection bias and representativeness
- Central limit theorem
- Point and interval estimates
- Standard error and precision
- Confidence interval interpretation
One-Sample Tests and Statistical Power
Statistical hypotheses, tails, z and t methods, p-values, errors, effect size and powerOne-sample inference tests a population claim using the distance between a sample estimate and a null value measured in standard-error units.
- Research and statistical hypotheses
- One-sided and two-sided tests
- One-sample z and t methods
- P-values and decision language
- Type I and Type II errors
- Effect size, sample size and power
Independent and Paired Sample Tests
Study design, paired differences, independent groups, variance checks and test interpretationTwo sets of measurements can arise from unrelated groups or from linked observations on the same units. That design distinction controls the analysis.
- Independent and paired designs
- Within-pair differences
- Paired t test
- Independent-samples t test
- Normality checks
- Levene's equality-of-variance test
Nonparametric Tests with Ranks
Parametric assumptions, rank-sum, Mann-Whitney U, signed-rank and ranked evidenceRank-based methods offer alternatives when small-sample distributional conditions for common t procedures are not credible or the measurement scale is ordinal. They do not erase study design.
- Parametric and nonparametric methods
- Distribution-free reasoning
- Wilcoxon rank-sum
- Mann-Whitney U
- Wilcoxon signed-rank
- Ties, zero differences and interpretation
Chi-Square Tests for Categorical Data
Contingency tables, expected frequencies, goodness of fit, association and cell contributionsChi-square methods compare observed categorical counts with counts expected under a null model. Goodness of fit uses one categorical variable and specified proportions, while association crosses two categorical variables and tests independence.
- Goodness-of-fit questions
- Tests of association
- Contingency-table margins
- Expected frequencies
- Cell contributions
- Degrees of freedom and pattern interpretation
Regression and Pearson Correlation
Scatterplots, explanatory and response variables, least squares, residuals and linear associationRegression and correlation summarise linear relationships between two quantitative variables. A scatterplot comes first because curvature and influential observations can make one coefficient misleading.
- Explanatory and response variables
- Scatterplot interpretation
- Least-squares fitted lines
- Slope and intercept
- Fitted values and residuals
- Pearson correlation and linear association
Screening Tests, Risk and Odds
Screening accuracy, cohort and case-control designs, relative risk and odds ratiosEpidemiological measures depend on the group used as denominator and the direction of study sampling. Screening accuracy conditions on true disease status, while predictive values condition on test results.
- Screening-test tables
- Sensitivity and specificity
- Predictive values
- Prospective and retrospective cohorts
- Case-control studies
- Relative risk and odds ratios
Discrete binomial and Poisson models lead into the normal distribution, sampling behaviour and confidence intervals.
The second half turns those foundations into inference: one-sample tests, independent and paired comparisons, rank-based alternatives, chi-square methods, simple linear regression, Pearson correlation, screening accuracy and epidemiological association measures.
How the course is organised
The Canvas sequence contains eleven modules.
Modules 1 to 4 build the descriptive and probability core assessed in the mid-semester summative task. Modules 5 to 9 develop estimation and hypothesis testing, including the choice between z and t procedures, independent and paired designs, parametric and nonparametric methods, and categorical inference. Modules 10 and 11 shift toward modelling relationships and interpreting common epidemiological measures.
Excel, SPSS and R support computation, but the course learning outcomes emphasise explaining methods and drawing conclusions under uncertainty.
What the assessment asks for
The published weights total 100%. Module Assignments carry 20%. Two on-campus summative assessments carry 15% and 25%. The project contributes the remaining 40% through a proposal, data file, poster and video presentation.
The assessment summary says non-multiple-choice responses must show working and unsupported final answers receive no marks. It also restricts outside apps and AI for module and summative tasks, so the current Canvas instructions control what may be used during each assessment.
Why method selection is the central skill
Many calculations look similar after software has produced a table.
The real distinction comes earlier: categorical or quantitative, count or measurement, one sample or two, independent or paired, normal-based or rank-based, cohort or case-control. A correct test chosen for the wrong design is still wrong.
This guide therefore makes every worked response state the research target, design, assumptions and denominator before interpreting a result.
How to study with this guide
Read the chapter map, reproduce each worked calculation, then attempt the practice cards without looking at the model answer.
Keep a method ledger that records the parameter, design, assumptions, formula or software quantity, and conclusion language for every procedure. Before the final summative test, mix questions from different modules so you must identify the method rather than recognise the chapter heading.
Worked example · free
Reverse a screening probability with Bayes' rule
- 1Define A as the risk factor and B as a positive result; list P(A), P(not A), P(B|A) and P(B|not A).
- 1Calculate P(A and B) = 0.30 x 0.80 = 0.24.
- 1Calculate the false-positive path 0.70 x 0.10 = 0.07 and total P(B) = 0.31.
- 1Divide 0.24 by 0.31 to obtain P(A|B) = 0.774.
- 1Interpret 77.4% among participants with a positive result, not among all participants.
Key terms
- Biostatistics
- The application of statistical methods to biological, medical and epidemiological data.
- Sample Statistic
- A numerical summary calculated from a sample and used to describe the sample or estimate a population parameter.
- Conditional Probability
- The probability of one event within the restricted group where another event is already known to have occurred.
- Binomial Distribution
- A probability model for the number of successes in a fixed set of independent binary trials with constant success probability.
- Poisson Distribution
- A probability model for the number of events in a stated exposure interval under a stable event rate.
- Z Score
- A dimensionless measure of how many standard deviations a value lies above or below a mean.
- Standard Error
- The standard deviation of a statistic's sampling distribution and a measure of sampling precision.
- Confidence Interval
- An interval produced by a method designed to capture a population parameter at a stated long-run rate.
- P Value
- The probability, assuming the null hypothesis, of the observed result or one more extreme in the tested direction.
- Statistical Power
- The probability that a test rejects the null hypothesis when the specified alternative is true.
- Paired Design
- A design in which measurements are linked within the same unit or a deliberately matched pair.
- Rank-Sum Test
- A nonparametric method that compares the pooled ranks of two independent groups.
- Chi-Square Test
- A categorical-data method that compares observed counts with counts expected under a stated null model.
- Residual
- The observed response minus the response fitted by a regression model.
MATH2209 FAQ
How is the course assessed?
The published structure totals 100%: Module Assignments carry 20%, the mid-semester and final summative assessments carry 15% and 25%, and the project carries 40% through a 5% proposal, 5% data file, 25% poster and 5% video. Check Canvas for current dates and instructions.
What does the final summative test cover?
The Semester 2 syllabus labels the final task Assessment 2 and states that it covers Modules 1 to 11. The assessment summary says it takes place on campus. Operational details such as time, room, permitted materials and software should be confirmed on Canvas because the syllabus marks them as changeable.
Which statistical method should I choose?
Start with the response variable and design. Counts may suggest binomial or Poisson models; one quantitative sample uses one-sample inference; two groups separate into independent or paired; violated parametric assumptions may point to rank methods; categorical counts use chi-square; cohort and case-control studies support different epidemiological measures.
Do I need to show working?
Yes for non-multiple-choice assessment responses. The assessment summary says answers without clearly shown working receive no marks. Write the formula or method, substitute values, keep intermediate quantities and units visible, and end with a sentence that interprets the result in the population and research context.
Can I use notes, software or a calculator?
The mid-semester assessment information says students should have notes, required software and a calculator available, but each task can carry its own restrictions. The current Canvas instruction is authoritative. Do not assume that permission for one assessment automatically applies to the final summative task or module assignments.
What dataset does the project require?
The project information asks for an analysis-ready dataset containing at least two numerical and two categorical variables, with at least two clearly defined research questions that align with methods taught in the course. The staged submissions move from proposal and data file to a scientific poster and three-minute video.
Can AI be used for assessment work?
The assessment summary restricts apps and AI for Module Assignments and both summative assessments, directing students to course materials, resources and instructions. Use this guide for revision and original practice outside restricted tasks, then follow the precise Canvas rule for the assessment you are completing.
Questions students ask in each MATH2209 chapter
- When should I report the median?
- Use the median when a quantitative distribution is strongly skewed or contains influential extreme values. It depends on order rather than every magnitude, so one unusual observation moves it less than the mean. Pair it with the interquartile range and still show the distribution. Chapter 01 · Data, Variables and Descriptive Statistics
- Why is sensitivity not the answer after a positive result?
- Sensitivity conditions on people who truly have the condition. A post-positive probability conditions on everyone who tested positive, including false positives. Because the denominator changes, Bayes' rule must include both positive paths rather than copy the sensitivity value. Chapter 02 · Probability Rules and Bayes
- How do I distinguish binomial from Poisson?
- Look for the mechanism. Binomial questions fix a number of binary trials and count successes. Poisson questions fix an exposure interval and count events arising at a rate. A count alone does not determine the model. Chapter 03 · Binomial and Poisson Models
- Why can a positive z score have a large probability?
- The z score only locates a boundary. The area to its left may be large, while the area to its right is small. Probability size depends on which region the question requests, not on the sign alone. Chapter 04 · Normal Distributions and Z Scores
- What makes an interval narrower?
- With other inputs fixed, a larger sample reduces standard error and narrows the interval. Lower confidence or lower variability also narrows it, but each change has a different interpretation. Precision should never be improved by silently changing the scientific target. Chapter 05 · Sampling, Estimation and Confidence Intervals
- What does a large p-value mean?
- It means the observed evidence is not sufficiently unusual under the null model for rejection at the chosen threshold. It does not prove the null is true, establish equality or show that the study had enough power to detect a meaningful effect. Chapter 06 · One-Sample Tests and Statistical Power
- How do I recognise paired data?
- Look for repeated measurements on the same unit or deliberate matching between units. Before-and-after readings, left and right measurements, and matched case-control pairs are linked. Two columns alone do not establish pairing. Chapter 07 · Independent and Paired Sample Tests
- Does nonparametric mean assumption-free?
- No. Rank procedures still require an appropriate design, independent observational units between groups or valid pairing, and interpretable ordering. They reduce reliance on a particular population distribution but do not make data quality and sampling assumptions disappear. Chapter 08 · Nonparametric Tests with Ranks
- When is goodness of fit appropriate?
- Use it when one categorical variable's observed counts are compared with specified expected proportions. If the question asks whether two categorical variables are related, construct a contingency table and use a test of association instead. Chapter 09 · Chi-Square Tests for Categorical Data
- Can correlation be near zero when variables are related?
- Yes. Pearson correlation measures linear association. A strong curved relationship can have positive and negative local slopes that cancel to near zero. Inspect the scatterplot before using correlation as a summary. Chapter 10 · Regression and Pearson Correlation
- Why do sensitivity and specificity use different denominators?
- Sensitivity evaluates detection within people who have disease, so its denominator is true positives plus false negatives. Specificity evaluates exclusion within people without disease, so its denominator is true negatives plus false positives. Chapter 11 · Screening Tests, Risk and Odds
How to prepare for the assessments
Build a selector sheet rather than a formula dump. For every method, record the target parameter, variable type, study design, assumptions, calculation or software output, and the exact language of the conclusion. Work through the course in three passes. First, secure Modules 1 to 4 by classifying variables, translating probability language and moving between binomial, Poisson and normal models.
Second, connect estimation to inference: standard error, confidence intervals, hypotheses, tails, z versus t, Type I and Type II errors, and power. Third, mix competing designs: independent versus paired, t versus ranks, goodness of fit versus association, regression versus correlation, and cohort risk versus case-control odds.
Recompute every worked number without looking, then alter one feature of the scenario and decide whether the method changes. The project offers a useful integration exercise: map each proposed research question to variables, design, plot, method, assumptions and reporting sentence before opening software.
During assessment preparation, follow the current Canvas restrictions, because the course summary limits outside apps and AI in graded tasks.
Your AI Statistics tutor for MATH2209
Stuck on a hard MATH2209 question? Sia is AskSia’s AI Statistics tutor — ask any MATH2209 Biostatistics question and get a clear, step-by-step explanation grounded in how the course is actually taught and assessed. Read this whole study guide free, then take your hardest questions to Sia.