QUANTITATIVE DATA ANALYSIS 2
See the mathematics behind the numbers.
An interactive laboratory where statistical concepts become things you can explore, manipulate and understand.
Don't just calculate it. See why it works. Drag the points on the right: the line, the residuals and r follow immediately.
r = 0.939 · R² = 0.881 · SSE = 517
Synthetic data: weekly news hours (X) and political knowledge (Y). Every coral square is a squared residual; the line minimises their total.
LEARNING PHILOSOPHY
The visualization is not an illustration beside the lesson. The visualization is the learning environment.
Most statistics teaching rests on text, static formulas and software output. Here the learner manipulates representations and watches the mathematical consequences in real time: an observation moves, the mean follows, deviations become squares, squares become variance.
Every value is computed from the actual data, from a single shared source. No fake numbers, no metric nudged for the sake of an animation. Sample and population variance, effect size and significance get separate names and separate explanations.
FEATURED LABORATORY
Variance → Regression → ANOVA
One dataset, three views. Variance measures the distance between observations and their mean; regression predicts with a line and measures what is left; ANOVA predicts with group means. All three ask the same question: how much variation remains after the model has made its prediction?
Open the labTopics
All topics →- DATA & MEASUREMENTMeasurement levelsModule 01
- DATA & MEASUREMENTIndex constructionModule 02 · 1 lab
- RELATIONSHIPSCrosstabsModule 03 · 1 lab
- RELATIONSHIPSLazarsfeld paradigmModule 04 · 1 lab
- INFERENCEp-valuesModule 05 · 1 lab
- RELATIONSHIPSCorrelationModule 06 · 1 lab
- INFERENCEt-testsModule 07 · 1 lab
- INFERENCEANOVAModule 08 · 2 labs
- DATA & MEASUREMENTDescriptive statistics · VarianceModule 09 · 1 lab
- Principal Component AnalysisModule 10
- LABSAll 20 laboratories →
EXPLORE
Open-ended experimentation, no account. Move points and parameters, switch representations, reset.
LEARN
Guided discovery: Predict → Experiment → Explain → Transfer. One conceptual question at a time, not a multiple-choice quiz.
TEACH
Projector view: large charts, collapsible panels, prepared demonstrations, reset to a known state. Classroom-ready.
Practise and go further
- Quick reviewThirty mixed questions across the modules; every wrong answer says which idea got mixed up.
- Go FurtherOptional depth: residuals, pairwise follow-ups, survey weights, p-value mathematics, reproducible SPSS.
- The Sopranos bonus labA whole analysis on a synthetic case dataset, from question to defensible conclusion.
Academic and educational use
This is an independently developed companion for the Quantitative Data Analysis 2 course. References to courses do not imply official institutional endorsement. SPSS remains the reference tool; the lab explains what its output means.
No accounts, no backend, no analytics. State stays in the browser. A future teaching pilot could study usability and conceptual understanding; learning effectiveness is a question, not a claim.