DATA1001 Foundations of Data Science
The Complete Exam Bible · S2 2026DATA1001 Overview
- University of Sydney
- Semester 2 2026
- First year undergraduate
- Closed book final, Exam Guide provided
DATA1001 Foundations of Data Science is a first year University of Sydney unit about reasoning from data. The eleven Canvas topics move from experimental design and graphical summaries through the Normal and linear models, chance, box models, sample surveys and hypothesis tests.
- Know the paper The published current format has 30 equal value multiple choice questions worth 75 percent and one extended data story worth 25 percent. It allows two hours plus ten minutes reading, is closed book, has no calculator and supplies the Exam Guide.
- Protect the largest assessment The official Semester 2 record gives the final exam 60 percent of the unit grade.
- Start with the heaviest topics In each current format sample, graphical summaries, numerical summaries and the linear model contribute 14 of 30 multiple choice questions, about 47 percent.
What you get in the full DATA1001 Exam Bible
- 7 chapters, 76 pages — every examinable topic, week by week
- 14 original practice questions in the Final exam format
- 7 worked examples, solved the way DATA1001 marks
- 60 computed figures, redrawn from the course’s own numbers
- Exam-morning plan for the Final exam (60%) — what to revise last, where marks are lost
Free on this page: assessment table, chapter map, one worked example, key terms, FAQ. The rest is in the full exam bible.
How DATA1001 is assessed
| Component | Weight | Format |
|---|---|---|
| Final exam | 60% | Written exam, two hours, formal exam period, statistical thinking with R output |
| Individual project report | 17% | Individual report due in Week 11 |
| Group project report | 8% | Group report due in Week 7 |
| Workshop contribution | 5% | Contribution during workshops |
| Weekly online quizzes | 4% | Online quizzes across teaching weeks |
| Individual project checkpoint | 3% | Individual checkpoint due in Week 9 |
| Group project presentation | 2% | Group presentation |
| Early feedback quiz | 1% | Early quiz due in Week 3 |
The official Semester 2 2026 record sets a two hour written final worth 60 percent. Published Semester 1 2026 exam information and the official November 2024 and June 2026 samples use 30 equal value multiple choice questions for 75 percent of the paper and one multipart data story for 25 percent, plus ten minutes reading, a supplied Exam Guide, closed book conditions and no calculator. It is reasonable to expect the Semester 2 paper to use the same frame, but this is an inference. The samples place most multiple choice weight on graphical summaries, numerical summaries and the linear model. Extended work asks for contextual interpretation and judgement across several sources.
What other students are asking about DATA1001
- Where can I find lecture notes or study guides for DATA1001 at USyd?
- The DATA1001 study guide in AskSia Library covers Foundations of Data Science in 7 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 DATA1001, and how much is it worth?
- Yes. In the offering this AskSia Library guide is built from, Final exam carries 60% of the DATA1001 mark; the full split is Final exam 60%, Individual project report 17%, Group project report 8%, Workshop contribution 5%, Weekly online quizzes 4%, Individual project checkpoint 3%, Group project presentation 2%, Early feedback quiz 1%.
- What does DATA1001 cover on exploratory data analysis?
- In AskSia Library's DATA1001 guide, the Exploratory Data Analysis chapter explains graphical summaries, numerical summaries, shape, centre, spread and outliers.
- What does DATA1001 cover on the normal distribution?
- In AskSia Library's DATA1001 guide, the The Normal Distribution chapter explains standard units, Normal areas, percentiles and measurement error.
- What does DATA1001 cover on regression?
- In AskSia Library's DATA1001 guide, the Regression chapter explains correlation, fitted lines, residuals, model fit and prediction.
Current DATA1001 dates
| Date | Item | Control |
|---|---|---|
| 9 to 15 November 2026 | STUVAC | University study vacation period. |
| 16 to 28 November 2026 | Examination period | University Semester 2 examination period. Check the personal timetable for the unit date and room. |
Dates come from the official University of Sydney 2026 wall calendar. The course materials were captured from the Semester 1 2026 Canvas site for the same eleven topics. The Semester 1 sitting was on 11 June 2026 at 17:00 and is not the next exam date. Students should use their personal examination timetable for the Semester 2 unit date and room. Confirm exact deadlines and submission settings in the live LMS.
What DATA1001 covers
Eleven Canvas topics form one path from design and description to modelling, sampling and decisions.
Study Design
Data types, observational studies, experiments, confounding, bias and samplingBefore calculation, study design determines what a result can support. Start by identifying the target population, sampled units, exposure or treatment, response and any grouping variables.
- Topic 1: Design Experiments
- Classify variables from their meaning
- Separate random sampling from random assignment
- Identify confounding and causal limits
- Diagnose selection, nonresponse and measurement bias
- Apply privacy, consent, fairness and transparency
Exploratory Data Analysis
Graphical summaries, numerical summaries, shape, centre, spread and outliersExploratory data analysis describes variables before a model or test is chosen. Begin with variable type because it determines which summaries and plots are meaningful. Categorical variables are described by counts and proportions, often with bar charts.
- Topic 2: Data and Graphical Summaries
- Topic 3: Numerical Summaries
- Match variable type to a legal display
- Compare shape, centre, spread and unusual values
- Choose resistant summaries under skew
- Read boxplot, histogram and scatterplot code
The Normal Distribution
Standard units, Normal areas, percentiles and measurement errorThe Normal model connects numerical summaries with probability. Its centre is the mean and its scale is the standard deviation. Standard units express a value as its signed distance from the mean in SDs: subtract the mean and divide by SD.
- Topic 4: Normal Model
- Standardise values with z scores
- Use the empirical rule for quick checks
- Read lower, upper and between areas
- Invert cumulative area with qnorm
- Separate chance error from systematic bias
Regression
Correlation, fitted lines, residuals, model fit and predictionRegression describes and predicts a quantitative response from a quantitative predictor with a fitted straight line. Correlation summarises the direction and strength of linear association.
- Topic 5: Linear Model
- Interpret correlation and its limitations
- Calculate slope, intercept and fitted values
- Read observed minus fitted residuals
- Interpret R squared and residual standard error
- Use residual plots and avoid extrapolation
Probability and the Box Model
Probability rules, binomial models, expected value, standard error and box modelsProbability supplies the chance model behind statistical inference. DATA1001 represents repeated chance processes with a box of tickets and a drawing rule.
- Topic 6: Understanding Chance
- Topic 7: Chance Variability
- Use union, intersection and conditional probability
- Check binomial conditions before calculating
- Build the course box model
- Scale expected value and standard error for sums and averages
Sampling Distributions
Parameters, statistics, sample surveys, standard error and confidence intervalsSampling distributions connect one observed sample to uncertainty about a population. A parameter is a fixed unknown population quantity. A statistic is computed from a sample and varies when the sampling process is repeated.
- Topic 8: Sample Surveys
- Topic 7: Chance Variability for sampling
- Separate parameters from statistics
- Use the central limit theorem with conditions
- Interpret standard error and confidence
- Distinguish precision from sampling bias
Hypothesis Testing
Hypotheses, test statistics, p values, mean tests and relationship testsHypothesis testing compares observed data with a null model through five linked moves: hypotheses, assumptions, test statistic, p-value and conclusion. The null hypothesis states a parameter value or lack of relationship.
- Topic 9: Hypothesis Testing
- Topic 10: Tests for Mean
- Topic 11: Tests for Relationship
- Use the five step testing frame
- Read t test and linear model output
- Calculate chi squared expected counts and interpret evidence
This guide follows that sequence while placing the published current exam format at the centre of revision. The official Semester 2 2026 record makes the final examination 60 percent of the unit grade.
The unit's Semester 1 2026 exam information and the November 2024 and June 2026 official samples show the current frame: 30 equal value multiple choice questions worth 75 percent of the paper and one extended data story worth 25 percent, completed in two hours after ten minutes reading.
The sitting is closed book, calculators and outside references are not allowed, and an Exam Guide with formulas and selected R commands is printed in the booklet. The Semester 2 paper is expected to follow this frame because the official record specifies the same two hour final, but that expectation is an inference rather than a published Semester 2 paper.
Revision should therefore centre on method recognition and interpretation. You need to identify a study design, choose summaries that match variable type and shape, standardise Normal values, read linear model coefficients and residuals, build a box model, distinguish standard deviation from standard error, and complete a hypothesis test with an in context conclusion.
The supplied papers repeatedly ask students to read R output and small pieces of code rather than operate R live. The ten output drills in this book cover linear models, t tests, proportion procedures, chi squared output, probability calls, tables and plots. The original mock then joins these skills in a fresh 30 question paper and one extended client story. Use the topic weights as a revision priority, not as a prediction.
Every topic remains examinable, and the extended response can connect design, plots, tests and judgement in one chain.
Worked example · free
Read a two sample t test in the paper's style
- 1Identify the comparison. The target is the population mean duration difference, A minus B.
- 1Read direction and size. The sample estimate is 6 minutes, with branch A higher.
- 1Read evidence. The p value is below 0.05 and the interval excludes zero, so equal population means are not supported at that level.
- 1Write context and restraint. The plausible mean difference is about 0.9 to 11.1 minutes. Importance depends on the council's operational threshold, and causation requires an appropriate design.
Key terms
- Observational study
- A study in which exposure is observed rather than randomly assigned, so confounding limits causal conclusions.
- Randomised experiment
- A study that assigns treatments by chance so comparison groups tend to balance confounders.
- Confounder
- A variable associated with both exposure and outcome that can distort their observed relationship.
- Median
- The middle ordered value, a resistant description of centre for skewed data.
- Interquartile range
- The distance from the first to the third quartile, describing the spread of the middle half.
- Standard deviation
- A measure of the typical distance of individual observations from their mean.
- Standard error
- The standard deviation of a statistic across repeated samples from the same model.
- Standard unit
- A value's signed distance from its mean measured in standard deviations.
- Correlation
- A unitless measure of the direction and strength of a linear association.
- Residual
- An observed response minus its fitted value, the vertical gap left by a model.
- Box model
- A representation of a chance process through tickets in a box and a specified drawing mechanism.
- Sampling distribution
- The distribution of a statistic across repeated samples generated in the same way.
- Confidence interval
- A data based range produced by a method with stated long run coverage.
- P value
- Under the null hypothesis, the probability of a test statistic at least as extreme as the observed one.
- Expected count
- The count predicted for a table cell by the null model, often independence.
- Statistical significance
- A decision that data are sufficiently incompatible with a null model at a chosen level.
DATA1001 FAQ
Is the final exam open book?
No. The published current format is closed book. The booklet supplies an Exam Guide with selected formulas and R commands, but students cannot bring reference materials. Revision should therefore separate lookup mechanics from concepts that must be recalled, especially design, p value meaning and the box model.
Can I use a calculator?
No. The published exam information says calculators are not permitted. Questions are designed around statistical reasoning and manageable arithmetic. Practise fractions, expected counts, simple standard errors and standardised distances by hand, and use the printed Exam Guide to locate formulas rather than trying to recreate every expression from memory.
How many questions and how long?
The current published format has 30 equal value multiple choice questions worth 75 percent of the paper and one extended question with several parts worth 25 percent. Writing time is two hours after ten minutes reading. A workable budget is 75 minutes, 30 minutes and 15 minutes for checking.
What is on the Exam Guide?
The guide carries numerical summary formulas, Normal and regression relationships, binomial and probability rules, the hypothesis test frame, and selected R commands for plots, distributions, models, simulation and t tests. It does not choose the method or supply the contextual interpretation that earns the reasoning marks.
Do I have to write R code?
The unit says the exam emphasises statistical thinking, application to new stories and interpretation of R output. It may show small pieces of R and ask what they produce or how to interpret them. There is no live R session. Learn to read commands for plots, probability calls, models, tables and tests.
How is the extended answer marked?
The official current format makes the extended question 25 percent of the paper, but the teaching copy does not state a mark split across its parts. Do not invent one. Answer every verb, use every source, write estimates with units and context, and separate statistical evidence from design limits and client judgement.
How do the two unit versions differ?
The captured final exam information page uses both unit codes in its title, while the materials supplied here establish the assessment for DATA1001 Foundations of Data Science. They do not provide enough verified detail for a broader comparison between the two units. Use the enrolled unit outline for any administrative difference.
How much is the final worth?
The official Semester 2 2026 assessment record assigns 60 percent of the unit grade to the final written examination. The remaining 40 percent is distributed across the individual and group projects, workshop contribution, online quizzes, a checkpoint, a presentation and an early feedback quiz.
What if I fail the final?
The official Semester 2 assessment record provided here states the final weight but does not state a separate hurdle or automatic fail rule. Do not assume one. Check the current unit outline and formal progression rules, and ask the unit coordinator if your personal result or special consideration situation needs an authoritative decision.
Questions students ask in each DATA1001 chapter
- When may I use causal language?
- Use causal language when treatment was randomly assigned and the comparison was otherwise conducted appropriately. Observational evidence can support association but leaves confounding open. A small p value does not upgrade an observational design into an experiment, so name the design before choosing words such as causes or changes. Chapter 01 · Study Design
- Mean or median on the exam?
- Use median when a distribution is skewed or contains influential outliers because it is resistant. Use mean for a roughly symmetric clean distribution when every value should contribute to centre. Explain the choice from the observed shape rather than treating either summary as universally superior. Chapter 02 · Exploratory Data Analysis
- When may I use a Normal model?
- Use it when the distribution is reasonably bell shaped for the task or when a justified sampling approximation applies. Computing a z score alone does not prove Normality. Inspect a histogram or other diagnostics and be cautious with strong skew, multiple modes or influential outliers. Chapter 03 · The Normal Distribution
- Does correlation establish causal effects?
- Correlation describes linear association and does not identify a causal effect. A strong value can arise from confounding, selection or reverse direction. Use the study design to decide whether causal language is available, even when the slope p value is very small. Chapter 04 · Regression
- When do probabilities add?
- Add probabilities for a union after subtracting any overlap counted twice. If events are mutually exclusive, the overlap is zero and simple addition works. Translate the word or into a union, but check whether both events can occur together before adding directly. Chapter 05 · Probability and the Box Model
- How do parameters and statistics differ?
- A parameter describes the population and is fixed although unknown. A statistic is computed from the sample and changes from sample to sample. Use notation and context carefully so a sample mean is not described as the population mean itself. Chapter 06 · Sampling Distributions
- What does a p value mean?
- Assume the null hypothesis is true, then ask how often a test statistic at least as extreme as the observed one would occur. It is not the probability that the null is true, the probability that results occurred by chance, or a measure of effect size. Chapter 07 · Hypothesis Testing
How to study for the exam
Start by protecting the assessments that occur before the final. The individual report carries 17 percent, the group report 8 percent, workshop contribution 5 percent, weekly quizzes 4 percent, the checkpoint 3 percent, presentation 2 percent and early quiz 1 percent. Use their feedback as evidence about recurring weaknesses rather than treating them as separate from exam study. For final preparation, take three passes.
First, read Exam Morning and write from memory what the Exam Guide carries and what it does not. Second, complete the ten R output drills, forcing every answer to name the parameter or relationship, the estimate, the evidence and the design limit. Third, sit the original mock in the published time structure, transfer all 30 choices to an answer sheet and complete the extended data story without a calculator.
Classify every error by Canvas topic. Begin repairs with Topics 2, 3 and 5 because they form 14 of 30 items in each current format sample, then cover the remaining topics rather than predicting omissions. In the last check, audit signs, tails, units, observed versus expected counts, SD versus SE, and association versus causation.
Rework any question you answered from a memorised phrase until you can explain why the alternatives are wrong in that particular story.