Quantitative UX (From Basics to Advanced)

Duration: 5 sessions × 2 hours (10 hours total)

Audience: UX designers and researchers 

Instructor: Mohsen Rafiei, Ph.D.

Format: Online, hands-on sessions using simple tools (Jamovi, JASP, R).

Registration Fee for all 5 sessions: $850

Email: Admin@Puxlab.com

The Quantitative UX (From Basics to Modeling) course takes UX researchers, and designers on a full learning journey, from the very basics of quantitative thinking to advanced analytical methods used in modern research. We start with simple, intuitive concepts like describing data and comparing designs, then gradually move into more sophisticated techniques such as regression, power analysis, and modeling. All sessions are hands-on: participants will receive curated resources, UX datasets, and guided exercises for practice. By the end of the course, researchers/designers will understand how to apply rigorous quantitative methods to their work, interpret data with confidence, and identify clear directions for deeper investigation and future learning in evidence-based UX research.

Session 1 - Descriptive & Parametric Statistics: Seeing the Story Behind the Numbers 

We start by turning data from something intimidating into something visual and meaningful. Designers already think in patterns and comparisons, that’s exactly what statistics is.

What you’ll learn

• Different data types (nominal, ordinal, interval, ratio) and how to handle each.

• Measures of center and spread: mean, median, standard deviation, variance.

• Visualizing UX data with histograms, boxplots, and bar charts.

• The logic of the normal distribution and why it matters for testing.

• How to compare designs using t-tests and one-way ANOVA.

• Understanding p-values, CI, and effect sizes in plain terms.

Applications

• Compare task completion times across two design versions.

• Evaluate whether a new layout really improves usability scores.

• Communicate results as visual stories rather than abstract numbers.

  
  

Session 2 - Non-Parametric Tests: Making Sense of Messy, Real-World Data

UX data rarely behaves perfectly. Ratings are skewed, sample sizes are small, and users don’t follow clean distributions. This session focuses on practical tools that still let you make solid decisions when the data breaks the rules.

What you’ll learn

• Why parametric assumptions fail and how to check them quickly.

• Median-based comparisons using Mann-Whitney U and Wilcoxon tests.

• Kruskal-Wallis and Friedman tests for comparing several designs.

• Chi-square for survey questions and categorical data.

• Bootstrapping ideas for small-sample reliability.

Applications

• Compare SUS or satisfaction ratings with fewer than 30 users.

• Test whether participants prefer one feature concept over another.

• Present statistically defensible findings even when your dataset isn’t “perfect.”

Session 3 - Bayesian Statistics: Updating Evidence Like a Designer

Designers work iteratively, testing - learning, improving - which makes Bayesian thinking a natural fit. Instead of asking “Did it work?” we ask “How sure are we now?”

What you’ll learn

• The Bayesian idea: starting with prior knowledge and updating it with new evidence.

• Credible intervals vs. confidence intervals, thinking in probabilities, not binaries.

• Simple Bayesian A/B testing for design iterations.

• How small samples can still produce useful insights with Bayesian logic.

• Communicating uncertainty honestly and visually.

Applications

• Ongoing design testing where results evolve as you collect data.

• Explaining evidence strength to stakeholders in a more intuitive way.

• Making confident, transparent decisions with limited user data.

Session 4 - Regression Analysis: Understanding What Drives User Outcomes

Regression helps you go beyond “is there a difference?” to “what actually causes it?” This is where you start modeling user experience.

What you’ll learn

• How regression works as a prediction and explanation tool.

• Linear regression for continuous outcomes (satisfaction, time, engagement).

• Logistic regression for binary outcomes (success/failure, yes/no choices).

• Reading coefficients as “influence strength” and direction.

• How to visualize relationships between variables clearly.

Applications

• Identify which usability factors predict satisfaction.

• Quantify how time-on-task and aesthetics together influence user trust.

• Prioritize design changes based on data-driven impact rather than intuition.

Session 5 - Modeling & Machine Learning: Advanced Methods for Smarter UX Decisions

The final session connects all the dots, how to use data not just to describe the past but to predict and understand future behavior. It’s about thinking like a designer who can partner effectively with data scientists.

What you’ll learn

• What modeling means in UX: simplified representations of user behavior.

• Multiple regression as a behavioral model (predicting satisfaction or errors).

• Clustering methods for user segmentation, turning raw metrics into personas.

• Simplifying long surveys using principal component analysis (PCA).

• Realistic, no-code introductions to how ML supports UX: churn prediction, feedback classification, personalization.

• When to trust models, when to question them, and how to keep results explainable.

Applications

• Find hidden user groups based on behavior or sentiment.

• Simplify large survey datasets into clear experience themes.

• Collaborate intelligently with analysts and data scientists on predictive UX projects.

By the end of this course, you’ll see that quantitative research isn’t reserved for statisticians, it’s a tool you can master to bring more clarity, confidence, and credibility to your work. You’ll leave not only knowing how to apply tests, analyze data, and interpret results, but also understanding how these techniques strengthen your design intuition rather than replace it. With these methods, you can plan and conduct studies that stand up to scrutiny, uncover deeper insights, and communicate findings that truly influence decisions. This course gives you the foundation to move beyond guesswork and toward a more rigorous, evidence-based practice, where your design choices are not just inspired, but supported by solid data and sound reasoning.