SOC613 Applied Analytical Statistics for Social Scientists
The Complete Exam Bible · SOC613 Overview
- Lingnan University
- Term 1, 2026-27
- Postgraduate
- Open-book final exam
- Three-credit required course
What this course asks you to do
SOC613 develops quantitative judgement for social research. Every method begins with a sociological question, a unit of analysis and an explicit measurement choice. Participation in lectures and class activities must reach 80%.
- Assessed by A 45% final exam, a 45% quantitative paper replication project and 10% participation.
- Exam conditions The final is an individual, two-hour, open-book assessment held in class.
- Eligibility condition Participation in lectures and class activities must reach 80%. Without approved justification, a shortfall may disqualify a student from the final assessment and produce a fail grade. This is a pass condition; confirm eligibility on the course LMS. The outline also uses the phrase minimum attendance requirement. A separate numeric threshold is unpublished.
- Core habit Match the research verb, measurement level, calculation and conclusion before reporting a result.
How SOC613 is assessed
| Component | Weight | Format |
|---|---|---|
| Class and tutorial participation | 10% | Preparation, discussion and applied analytical work |
| Quantitative Paper Replication Project | 45% | Group report of 1,500 to 2,000 words with fully reproducible code |
| Final exam | 45% | Individual, open-book, in class, two hours |
The numeric assessment rows total 100%. Participation in lectures and class activities must reach 80%. Without valid justification, a shortfall may result in disqualification from the final assessment and a fail grade. This is a pass condition; confirm eligibility on the course LMS. The outline describes a minimum attendance requirement. A separate numeric threshold is unpublished.
Current SOC613 dates
| Date | Item | Control |
|---|---|---|
| 2026-12-09 | Final exam | |
| 2026-12-23 | Replication project |
Dates are as published in The official course outline publishes both dates for Term 1, 2026-27.. Confirm exact deadlines and submission settings in the live LMS.
What SOC613 covers
Five chapters follow the detailed captured sequence from research questions and measurement through reproducible data work, descriptive statistics and categorical association.
Research Questions and Causal Claims
descriptive, explanatory, predictive and causal questions · units and variables · association, temporal order and confoundingQuantitative analysis begins by deciding what kind of claim the evidence must support. This chapter separates descriptive, explanatory, predictive and causal questions, then connects each to a suitable comparison and conclusion.
- Descriptive, explanatory, predictive and causal questions
- Units of observation and analysis
- Dependent and independent variables
- Hypotheses and observable implications
- Association and temporal order
- Confounding and counterfactual reasoning
Variables, Concepts and Measurement
concepts, indicators and operational definitions · nominal, ordinal and interval-ratio levels · binary meansMeasurement connects an abstract sociological concept to observable evidence. This chapter separates concepts, indicators and recorded variables, then asks whether an operational definition captures the intended construct.
- Concepts, indicators and variables
- Operational definitions and comparability
- Validity and reliability
- Nominal, ordinal and interval-ratio levels
- Binary indicators and proportions
- Recoding and composite measures
Data Management and Reproducible Research
rows, variables and codebooks · recoding and missing values · auditable scripts and verificationA reproducible analysis preserves the received data and records every transformation in code. This chapter starts with rows, columns, cells, units and codebooks, then moves through recoding, constructed variables and missing-value handling.
- Rows, columns and units of observation
- Codebooks and raw frequencies
- Recoding without overwriting
- Constructed variables and scale direction
- Missing values and analytic samples
- Scripts, projects and verification checks
Central Tendency, Dispersion and Visualisation
mode, median and mean · range, interquartile range, variance and standard deviation · shape and outliersDescriptive analysis asks what is typical, how much observations vary, what shape the distribution takes and whether unusual cases require investigation. This chapter links mode, median and mean to measurement level and sensitivity.
- Mode, median and mean
- Binary means as proportions
- Range and interquartile range
- Variance and standard deviation
- Skewness and distribution shape
- Outliers and defensible visualisations
Cross-Tabulation and Chi-Square
conditional percentages · expected frequencies and Pearson chi-square · degrees of freedom, Cramer's V and lambdaCross-tabulation describes how two categorical variables occur together. This chapter identifies cells, marginals and table dimensions, then calculates percentages within the independent variable.
- Cells, marginals and table dimensions
- Conditional percentage direction
- Existence, strength and direction
- Expected frequencies under independence
- Pearson chi-square and degrees of freedom
- Cramer's V, lambda and substantive importance
Without approved justification, falling below that level may lead to disqualification from the final assessment and a fail grade. This is a pass condition; confirm eligibility on the course LMS. The outline calls this a minimum attendance requirement.
A separate numeric threshold is unpublished.
Research logic and measurement
The opening chapters distinguish descriptive, explanatory, predictive and causal questions, then connect concepts to indicators and recorded variables. Causal language requires association, temporal order and a defensible response to confounding.
Nominal, ordinal and interval-ratio levels determine which summaries and arithmetic have substantive meaning. A number stored in a column is not automatically quantitative, and a zero-one mean must be translated into the proportion assigned one.
Data preparation and description
Reproducible work starts with an unchanged source file and a codebook.
Every transformation receives a new name, an explicit rule and a verification table. Missing-value codes are handled before comparisons or summaries. Descriptive analysis asks what is typical, how widely cases vary, what shape the distribution takes and whether unusual values need investigation.
Center, spread and visualisation answer different questions and should be reported together.
Categorical relationships
Cross-tabulation classifies cases jointly by two variables. Percentages are calculated within the independent variable, expected counts express independence and Pearson's chi-square combines observed-to-expected discrepancies. Cramer's V describes strength.
Interpretation returns to conditional percentages and separates statistical evidence, practical importance and causal explanation.
Assessment preparation
The replication project reconstructs a published quantitative sociology analysis from the authors' data and requires code another analyst can run. The final is individual, open book, in class and two hours long.
Its published emphasis is application, causal reasoning, analytical rigour and clear interpretation. Because no sections or question types are published in the captured materials, revision should focus on choosing methods, checking assumptions, calculating transparently and writing conclusions whose strength matches the design.
Worked example · free
Compute and interpret one expected count
- 1Write the expected-frequency formula and identify both marginals.
- 1Calculate 180 times 240 divided by 600.
- 1State the result, 72 cases.
- 1Interpret it as the count predicted under independence.
Key terms
- Unit of Analysis
- The entity about which a study draws its conclusion, such as a person, household or organisation.
- Dependent Variable
- The outcome a study seeks to describe, explain or predict.
- Confounder
- A variable related to both a proposed cause and an outcome that can create a misleading association.
- Operational Definition
- The rule connecting an abstract concept to observable indicators and recorded values.
- Nominal Variable
- A categorical variable whose values differ without carrying an inherent rank order.
- Ordinal Variable
- A categorical variable with ordered values but no assumed equal distance between adjacent categories.
- Codebook
- Documentation that explains variable names, labels, response codes and missing-value conventions.
- Reproducibility
- The ability to recreate a dataset and result from the stated source files and analysis code.
- Interquartile Range
- The distance from the first quartile to the third quartile, covering the middle half of ordered observations.
- Expected Frequency
- A cell count implied by statistical independence while observed row and column marginals remain fixed.
- Chi-Square Test
- A test that aggregates observed-to-expected discrepancies across a categorical contingency table.
- Cramer's V
- A zero-to-one measure of association strength derived from chi-square, sample size and table dimensions.
SOC613 FAQ
How is this course assessed?
Participation contributes 10%, the Quantitative Paper Replication Project contributes 45%, and the final exam contributes 45%. The numeric weights total 100%. Attendance also has a separate minimum condition that can affect eligibility for the final assessment and the overall course result.
What is published about the final exam?
The official outline describes an individual, open-book, in-class final lasting two hours on 9 December 2026. It emphasises applying multivariate methods and causal-inference principles with sociological reasoning and clear interpretation. Sections and question types are not published in the captured materials.
What does the attendance rule require?
Participation across lectures and class activities must reach 80%. Without valid justification, a lower rate may disqualify a student from the final assessment and lead to a fail grade. This is a pass condition; confirm eligibility on the course LMS. The outline describes a minimum attendance requirement. A separate numeric threshold is unpublished.
What must the replication project contain?
The project reconstructs the core findings of a selected quantitative sociology article using the same dataset. Groups submit a report of 1,500 to 2,000 words and fully reproducible code, explaining the research question, modelling choices, reconstructed results and any differences from the published findings.
Why does measurement level matter before analysis?
Measurement level determines which summaries and operations have substantive meaning. Nominal values support classification, ordinal values add order, and interval-ratio values add magnitude. Software can average numeric codes even when those codes only name categories, so the analyst must make the decision.
How should missing values be handled in R?
Start from the codebook, convert documented nonresponse codes explicitly and preserve the original variable. Report valid counts, create transformed fields with new names and verify them through tables or summaries. Avoid silently deleting different sets of observations across models.
When should the median be preferred to the mean?
The median is useful for ordered or quantitative data when a distribution is skewed or contains plausible extreme values. Pair it with the interquartile range and a graph. The mean remains informative when its sensitivity is acknowledged and the variable supports arithmetic.
What does a significant chi-square result establish?
It provides evidence that the categorical variables are not statistically independent under the test assumptions. It does not establish causation or practical importance. Interpret the conditional percentages, report effect size and explain which cells contribute to the observed pattern.
Questions students ask in each SOC613 chapter
- How do I distinguish prediction from causation?
- Prediction asks whether available information anticipates an outcome, while causation asks whether changing one factor would change that outcome. A variable can predict well without being a cause. State which question is being answered before choosing the method or interpretation. Chapter 01 · Research Questions and Causal Claims
- Can I average a variable stored as numbers?
- Only when the numbers represent meaningful magnitudes or a justified approximation. Category codes may be nominal or ordinal even though software stores them numerically. Read the codebook and explain the measurement level before selecting a mean. Chapter 02 · Variables, Concepts and Measurement
- Why should raw data remain unchanged?
- An unchanged received file provides the fixed starting point for verification. Every cleaning decision can then be rerun, inspected and reversed through code. Manual edits erase the audit trail and make disagreements impossible to locate reliably. Chapter 03 · Data Management and Reproducible Research
- Which measures of center fit each level?
- The mode is available at every level, the median requires order, and the mean requires meaningful approximately equal intervals. Choose from substantive measurement first, then use distribution shape and the reporting goal to decide among valid options. Chapter 04 · Central Tendency, Dispersion and Visualisation
- Which direction should percentages run in a table?
- Calculate percentages within categories of the independent variable. If it appears in columns, each column should total one hundred per cent and comparisons run across rows. State the denominator so the direction is visible to the reader. Chapter 05 · Cross-Tabulation and Chi-Square
How to study for the exam
Build one analysis notebook that follows the course sequence. For every practice dataset, begin with a four-line statement of the research question, unit, outcome and explanatory variable. Add operational definitions and measurement levels before writing code. Preserve the raw file, document missing codes, create transformed variables under new names and insert verification tables immediately after each change.
Describe every variable before modelling it, including valid N, center, spread, shape and unusual observations. When examining a relationship, write conditional percentages first, then compute the appropriate test and effect size. Finish each session with a paragraph that reports the result in sociological language and limits the conclusion to what the design supports.
For the open-book exam, turn this notebook into a short index rather than a pile of notes: one page for question types and causal claims, one for measurement, one for reproducibility, one for distribution summaries and one for categorical relationships. Practise finding a rule quickly, then close the notes and explain why it applies.
For the project, keep a decision log and rerun the complete script from the unchanged source file before submission.
Your AI Statistics tutor for SOC613
Stuck on a hard SOC613 question? Sia is AskSia’s AI Statistics tutor — ask any SOC613 Applied Analytical Statistics for Social Scientists 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.
