SOST10031 Principles of Data Science
SOST10031 Overview
- 20 credit module
- The University of Manchester
- Semester 1 2026/27
- First year undergraduate
- Assessed by individual portfolio
SOST10031 Principles of Data Science is a first-year, 20-credit module from the Department of Social Statistics at The University of Manchester, taught in Semester 1 2026/27. It introduces social data science through one running question: how do different measures portray gender inequality across countries?
- Assessed by One individual data science portfolio of up to 2,000 words.
- Hardest step Writing claims that checked evidence can actually support.
- Programming load Small guided changes to supplied R code in Quarto.
- Habit to build Check meaning, source, timing and coverage before interpreting.
What you get in the full SOST10031 Study & Assessment Guide
- 10 chapters, 53 pages — every assessed topic, week by week
- 26 original practice questions, each with a model answer
- 10 worked examples, solved the way SOST10031 marks
- 15 computed figures, redrawn from the course’s own numbers
- Before-you-submit checklist for the Data science portfolio (100%) — 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 SOST10031 is assessed
| Component | Weight | Format |
|---|---|---|
| Data science portfolio | 100% | Individual portfolio of up to 2,000 words: a guided analysis of the supplied UNDP data in Quarto and your own investigation proposal. Due Friday 11 December 2026 at 17:00 through Canvas Assignments. |
| Formative portfolio plan | 0% | A plan of up to 250 words excluding the reference that does not count towards the final mark. Due Friday 30 October 2026 at 17:00, with feedback after reading week. |
What other students are asking about SOST10031
- Where can I find lecture notes or study guides for SOST10031 at manchester?
- The SOST10031 study guide in AskSia Library covers Principles of Data Science in 10 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 SOST10031?
- No final exam is listed in the offering this AskSia Library guide is built from; SOST10031 is assessed by Data science portfolio 100%.
- What does SOST10031 cover on computational thinking and r?
- In AskSia Library's SOST10031 guide, the Computational Thinking and R chapter explains timetable week 3. Decomposition, abstraction, sequencing, precision and checking, pseudocode, reading supplied R code, and errors versus analytical mistakes.
- What does SOST10031 cover on data sources, provenance and data quality?
- In AskSia Library's SOST10031 guide, the Data Sources, Provenance and Data Quality chapter explains timetable week 4. Fitness for purpose, meaning, provenance, timing, coverage, completeness, comparability and missing values recorded as NA.
- What does SOST10031 cover on exploratory data analysis?
- In AskSia Library's SOST10031 guide, the Exploratory Data Analysis chapter explains timetable week 5. Variable types, first displays, distributions, centre and spread, mean versus median, boxplots, scatterplots and association.
Current SOST10031 dates
| Date | Item | Control |
|---|---|---|
| 30 October 2026 | Formative portfolio plan due | Submission by 17:00; the plan carries no weight in the final mark. |
| 11 December 2026 | Data science portfolio due | Submission by 17:00 through the Assignments section of Canvas. |
Dates are as published in the Semester 1 2026/27 Canvas assessment and feedback page. Confirm exact deadlines and submission settings in the live LMS.
What SOST10031 covers
Ten chapters follow the 2026 timetable from the data science lifecycle and R to data quality, exploration, linkage, reproducibility, charts, ethics, communication and the portfolio.
Social Data Science: From Question to Conclusion
Timetable week 2. The question, data, analysis and communication loop, measured outcomes and reported attitudes, and the UNDP teaching datasets.Social Data Science: From Question to Conclusion sets up the habits the rest of SOST10031 relies on.
- The question, data, analysis and communication loop
- Social questions, statistical thinking and computing tools
- Measured outcomes in the Gender Inequality Index file
- Attitudes reported in the GSNI teaching file
- Labour force participation and its age 15 and over base
- Percentage points compared with percentage change
Computational Thinking and R
Timetable week 3. Decomposition, abstraction, sequencing, precision and checking, pseudocode, reading supplied R code, and errors versus analytical mistakes.Computational Thinking and R covers how SOST10031 turns a question into ordered, checkable instructions.
- Five habits of computational thinking
- Ordering a data task into checkable steps
- Pseudocode as a plan written for people
- Reading assignment, the pipe and filter aloud
- Predicting output before running a line
- R errors compared with analytical mistakes
Data Sources, Provenance and Data Quality
Timetable week 4. Fitness for purpose, meaning, provenance, timing, coverage, completeness, comparability and missing values recorded as NA.Data Sources, Provenance and Data Quality asks whether data suit the conclusion you want to draw, even when the code is correct.
- Quality as fitness for purpose
- Data sources compared with teaching datasets
- Provenance from original record to conclusion
- Publication, reference and observation dates
- Coverage compared with completeness
- Missing values recorded as NA, not zero
Exploratory Data Analysis
Timetable week 5. Variable types, first displays, distributions, centre and spread, mean versus median, boxplots, scatterplots and association.Exploratory Data Analysis is the week in which you learn to look at a dataset before deciding what it shows.
- What one row represents in a dataset
- Categorical variables, counts and measurements
- Matching a question to a first display
- Centre, spread, shape and unusual values
- How the mean and the median respond to an extreme value
- Building and reading a boxplot
Databases and Data Linkage
Timetable week 6. Joining country-level files, checking country names, unmatched records, failed matches versus missing values and the comparison group.Databases and Data Linkage brings together the two teaching files that the module has so far used separately.
- Linking the GII and GSNI teaching files by country
- Checking country names before a join
- Counting matched and unmatched rows
- Failed match versus missing indicator value
- Which countries form a complete pair
- Different reference periods in linked data
Reproducibility and Automation
Timetable week 8. RStudio projects, Quarto rendering, supplied workflows, automated checks with stopifnot, documentation and checked AI assistance.Reproducibility and Automation explains the working habits that let another person rerun your analysis and reach the same results.
- Project files and the working directory
- How R, RStudio and Quarto divide the work
- What rendering a Quarto document does
- Installing packages once and loading them each session
- Automated checks with stopifnot
- Run, Source and Source with Echo
Data Visualisation and Chart Critique
Timetable week 9. Matching a question to a chart, axes and scales, overlapping categories, bin width, clarity and accessibility.Data Visualisation and Chart Critique follows timetable week 9, whose seminar focuses on critiquing charts and improving their clarity and accessibility.
- Overlapping categories that cannot be summed
- Choosing a first display for a question
- Axis ranges and truncated scales
- Histogram bin width and hidden shape
- Direct value labels and informative captions
- Accessible colour and readable labels
Ethics, Representation and Responsible Analysis
Timetable week 10. Bias, representation, privacy, whose experiences the statistics capture and the consequences of an analysis.Ethics, Representation and Responsible Analysis covers timetable week 10, where the seminar discusses bias, representation, privacy and the consequences of analysis. In this module those themes are practical rather than abstract.
- Whose experiences the statistics represent
- Missing GII values and the countries left out
- Survey waves, coverage and the cost of closer dates
- Bias as a classification of responses
- Attitudes versus behaviour
- Country-level patterns and individual lives
Communicating Findings
Timetable week 11. Claims that fit the evidence, qualifications that matter, data stories for a non-specialist reader and peer feedback.Communicating Findings follows timetable week 11, when the seminar is spent giving and receiving feedback on a draft data story. The chapter treats communication as a sequence of careful rewrites.
- Claim strength from overstated to defensible
- Naming the measured indicator
- Attaching publisher and observation period
- Qualifications that are specific and material
- Percentage points versus percentage change
- Data stories for a general-interest reader
Data Science in Practice and Portfolio Decisions
Timetable week 12. The guided analysis, the investigation proposal, the formative plan, sources, Copilot reflection and submission.Data Science in Practice and Portfolio Decisions brings the module together in timetable week 12, when the seminar turns to portfolio decisions.
- Two official descriptions of the portfolio
- The 2,000-word budget and its exclusions
- Six published marking criteria
- The five elements of the formative plan
- Key dates from first lecture to deadline
- Checked use of Copilot in Part A
Two curated UNDP teaching files carry that question through the semester.
The Gender Inequality Index file holds 195 countries and territories with outcome indicators for reproductive health, secondary education, parliamentary seats and labour force participation, while the Gender Social Norms Index file holds 54 countries from the 2017 to 2022 World Values Survey wave and summarises reported attitudes towards gender roles.
The ten teaching weeks move from the data science lifecycle and computational thinking in R to data sources and data quality, exploratory analysis, databases and data linkage, reproducibility and automation, visualisation, ethics and responsibility, communicating findings, and data science in practice.
Lectures introduce each idea with short demonstrations in R; discussion-based seminars ask you to explain what code does, critique charts and judge whether a claim fits the evidence; guided independent exercises build programming confidence without assuming any previous experience.
Reproducibility, uncertainty, representation and ethical responsibility run through every week rather than sitting in one lecture, and the University-provided Copilot is treated as a tool whose suggestions you must check.
Assessment rests on a single individual Data Science Portfolio of up to 2,000 words.
You run a supplied Quarto analysis, apply one small guided change, verify the output, assess a claim and explain a finding for a non-specialist reader, then propose an investigation of your own using one or both teaching files. This guide follows the timetable chapter by chapter, works original examples with every number shown, and closes with portfolio-shaped practice and a submission-day checklist.
Worked example · free
Say which countries an average GII value actually describes
- 1Explain the NA: when any value in the column is missing, the mean cannot be computed until you decide how to treat those rows.
- 1Count the rows used: the file reports 23 missing GII values, so 195 minus 23 leaves 172 rows with a value.
- 1Express coverage: 172 divided by 195 is 0.882, so the average draws on 88.2% of the rows.
- 1Report the figure with its group: about 0.329 is the mean GII among the 172 countries and territories that have an available value.
- 1Add the check: removing NA does not repair the gaps or prove the remaining rows are typical, so inspect which places are missing, such as Andorra, before generalising.
Key terms
- Social data science
- Putting data to work on questions about people and society, combining social questions, statistical thinking and computing tools.
- Labour force participation rate
- The share of those aged 15 or older who either hold a job or are unemployed while seeking and available for work.
- Percentage point
- The unit for a difference between two percentages, distinct from a relative percentage change.
- Gender Inequality Index
- A UNDP index from 0 to 1 combining reproductive health, empowerment and labour market indicators, where higher values mean greater measured inequality.
- Gender Social Norms Index
- A UNDP measure built from survey answers that records the share of respondents showing at least one measured bias against gender equality.
- Data provenance
- The recorded history of a value, from its original measurement through each preparation step to the dataset in use.
- Fitness for purpose
- The standard that judges data quality by whether evidence suits its intended use and its limitations are understood.
- Missing value
- An entry recorded as NA because the value is unavailable in the dataset, which is different from an observed zero.
- Reference-year series
- The set of indicator columns selected for a given year, whose underlying observations may come from earlier years.
- Data linkage
- Combining two datasets by matching records, such as country names, and then investigating any records that fail to match.
- Quarto document
- A source file that combines written explanation, R code, tables and figures and renders them into a finished document.
- Reproducible analysis
- An analysis that another person can rerun with the same code and data to obtain the results reported.
SOST10031 FAQ
How difficult is Principles of Data Science?
The whole module mark comes from one portfolio of up to 2,000 words, so the full weight sits on a single submission. No previous programming experience is assumed and you are not asked to write R code from scratch. The demanding part is judgement: checking output, matching each claim to its evidence, finding a relevant source yourself for each part and evaluating Copilot suggestions.
How is the module assessed?
An individual data science portfolio carries 100% of the mark and is due on Friday 11 December 2026 at 17:00 through Canvas Assignments. A 250-word formative portfolio plan, due at 17:00 on Friday 30 October 2026, counts for nothing but earns feedback after reading week. The final assessment brief sets out the marking criteria and word-counting rules.
What goes into the portfolio?
Canvas and the draft outline describe Part A, a guided analysis of about 1,400 words in which you run a supplied Quarto file, make a guided change, check results, judge a claim and explain one finding plainly, and Part B, an investigation proposal of about 600 words. The module outline names the same work as annotated analysis, critical reflection and data story.
Do I need programming experience before starting?
No. R is introduced gradually through demonstrations, supplied examples and guided exercises, starting with installing R and RStudio and rendering a supplied Quarto file. Later weeks ask you to read, run and slightly adapt supplied code. When something breaks, bring the code, the complete error message and what you expected to happen to a seminar.
Can I use AI tools for the portfolio?
The module builds AI use into the portfolio through the University-provided Microsoft 365 Copilot, so no paid subscription is needed. You draft your own interpretation, ask Copilot to critique it, test its suggestions against your results and the data documentation, and explain what you accepted or rejected. Its contribution must be acknowledged and any code or claim it offers checked.
What counts towards the 2,000-word limit?
Your written explanation counts. R code, statistical outputs including tables and graphs, and the reference list are excluded from the word count. The suggested lengths of about 1,400 words for Part A and 600 for Part B are planning guides. The resit keeps the same portfolio format and the same 2,000-word maximum.
How should I plan a teaching week?
The weekly Canvas pages suggest one hour of lecture, one to two hours of reading and seminar preparation, one hour of seminar, two to three hours of guided R activities and thirty minutes of review. Bring incomplete answers and questions to seminars, and start the formative plan before reading week so its feedback arrives in time for Part B.
Questions students ask in each SOST10031 chapter
- Does a 24% participation rate mean women are a quarter of the workforce?
- No. Only women aged 15 and over sit in that rate's base. The female share of everyone in the labour force uses a different denominator, all workers of any sex, and a single participation column cannot provide it. Chapter 01 · Social Data Science: From Question to Conclusion
- Do I need to write R code from scratch?
- No. Students read supplied code, explain each instruction in ordinary language, predict its output and make small guided changes. The portfolio likewise works from a supplied Quarto analysis rather than asking for an analysis built independently. Chapter 02 · Computational Thinking and R
- Why can't I just replace NA with zero?
- Zero is an observed result, while NA marks information that is unavailable. Filling gaps with zeros invents measurements and can distort averages, rankings, charts and relationships, usually by pulling summary figures towards values that were never recorded. Chapter 03 · Data Sources, Provenance and Data Quality
- Should I report the mean or the median?
- Report the summary that represents the typical case and say why you chose it. When a distribution is skewed or contains a very large value, the median usually describes most observations better, while the mean is pulled toward the tail. Quoting both lets a reader see the effect of that tail directly. Chapter 04 · Exploratory Data Analysis
- Why do some countries disappear when two files are joined?
- A country drops out if it appears in only one file, if its name is written differently in each file, or if a needed value is recorded as NA. Listing the unmatched records and giving each a reason shows which of these happened and whether any loss can be repaired. Chapter 05 · Databases and Data Linkage
- Why must I open the .Rproj file instead of the .qmd file?
- Opening the project sets the working directory to the project folder, so relative paths such as data/ resolve on any computer. Opening a document directly can leave R looking in a different folder, which produces errors about files that seem to be missing even though they are present. Chapter 06 · Reproducibility and Automation
- Why can the GSNI dimension percentages not be added together?
- Each dimension share counts respondents classified as biased on that dimension, and one respondent can qualify on several. Adding the shares double counts those people, so the total describes no real group. Only the at-least-one and no-measured-bias figures form a complementary pair that reaches 100%. Chapter 07 · Data Visualisation and Chart Critique
- Does a GSNI bias label mean a person is prejudiced?
- No. UNDP applies the label to a survey response that meets one of its predefined rules. It summarises answers consistently across countries and is not a judgement of character, nor evidence of how the respondent behaves in daily life. Chapter 08 · Ethics, Representation and Responsible Analysis
- How do I explain a statistical finding to someone without training?
- Name the group and measure in everyday words, express a rate as so many in every 100 of that group, give comparisons in percentage points, and finish with the one caution most likely to prevent a misreading, such as what the indicator leaves out. Chapter 09 · Communicating Findings
- Which parts of the semester's work feed the one marked task?
- Everything converges on a single individual Data Science Portfolio of 2,000 words carrying the whole module mark. Seminar notes, guided R exercises and the 250-word formative plan prepare it; the plan returns written feedback after reading week while carrying no weight, so it acts as a rehearsal for Part B. Chapter 10 · Data Science in Practice and Portfolio Decisions
How to prepare for the assessments
Treat SOST10031 as practice in careful reasoning with real data rather than as a programming course. Each week, read the notebook alongside the lecture slides and predict what every supplied line of R will produce before you run it; the habit of predicting and then checking is the skill the portfolio rewards. Keep a running decision log with three columns: the choice you made, the check you ran and the conclusion it allowed.
Return to the two UNDP files often, because familiarity with their columns, footnotes and missing values makes later linkage and charting weeks far easier. Draft the formative plan early, using the five elements the brief lists, and treat its feedback as the first draft of Part B.
From week 9 onwards, practise turning each chart you meet into two sentences for a reader who has never studied statistics, with one caution about what the chart cannot show. Finally, render your Quarto file regularly from a clean start, so that reproducibility problems surface in November rather than on submission day.
Your AI Computer Science tutor for SOST10031
Stuck on a hard SOST10031 question? Sia is AskSia’s AI Computer Science tutor — ask any SOST10031 Principles of Data Science 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.