The Stratum blog

Statistics, visualized — one post at a time

How-tos, worked case studies, plain-language statistics explainers, and product news. The written companion to our video series — new posts every week.

🔖 Subscribe via RSS  ·  New to Stratum? Start the 25-lesson tutorial →

Latest posts

How-To

Open an SPSS .sav File on a Mac (No SPSS Licence)

Read variable and value labels straight off a .sav — no SPSS, no CSV round-trip.

Statistics Explained

Cp, Cpk, Pp, Ppk: Process Capability Explained

The four capability indices in plain language — and the assumption they all share.

How-To

Welch's ANOVA: When Variances Aren't Equal

Classical ANOVA crowns a false winner (p=0.0005); Welch says it's noise (p=0.42).

Statistics Explained

Type I, II, and III Sums of Squares

On unbalanced data the same effect gets three different F-tests. Which is right?

Statistics Explained

Effect Size: Cohen's d, Hedges' g, and Eta Squared

A p-value says whether an effect exists, not how big. Report the size too.

How-To

ANCOVA: Comparing Groups While Adjusting for a Covariate

Compare group means the fair way — hold a covariate constant, straight from menus.

Comparisons

Stratum vs Jamovi & JASP

Two free R-backed apps versus a paid native one — an honest head-to-head.

Statistics Explained

What Is Correspondence Analysis?

PCA for categorical data: turn a big cross-tab into one readable map.

Product Updates

What's New in Stratum 1.3

Polynomial regression terms, a solver validated against NIST, coefficient intervals and Effect Tests, ANOVA means and effect sizes, and Type III sums of squares.

Statistics Explained

Simpson's Paradox in the Wild

The best-cut diamonds look cheapest — until you find the confounder hiding in plain sight.

Statistics Explained

When to Log-Transform Skewed Data

When the log transform earns its place — and why it isn't just cosmetic.

Statistics Explained

Reading a PCA You Didn't Write

Name each component from its loadings — and let PCA explain a real mystery.

How-To

Multiple Regression That Isn't Fooled by Confounders

Add one variable and watch the R² climb — and the story flip.

Statistics Explained

Decision Trees on Real Data

Let 54,000 diamonds carve their own rules — and confirm the finding from a new angle.

How-To

How to Make a Q-Q Plot (Mac & Windows)

Check whether your data is normal — read the line, spot skew and tails.

Comparisons

Stratum vs SPSS

Where SPSS's established workflows fit, and where a lighter native app wins on cost.

How-To

How to Build a Decision Tree (Mac & Windows)

An interpretable model: grow it, prune it, read importance, and predict.

Statistics Explained

Violin vs Box Plot: Which Should You Use?

Both compare groups, but one shows summaries and the other shows shape.

How-To

How to Make a Violin Plot (Mac & Windows)

Compare groups by full distribution shape — and spot bimodality a box plot hides.

Comparisons

Stratum vs GraphPad Prism

Where Prism's biostatistics focus shines, and where a broader native app fits.

How-To

How to See Every Variable Pair at Once

See every variable pair at once with the Scatter chart's Matrix mode.

Statistics Explained

How to Read a Correlation Matrix

Read every pair's relationship at once, with the significant ones highlighted.

Statistics Explained

Reading Regression Diagnostics

R², VIF, Cook's distance and residual plots — whether to believe the model.

How-To

How to Run a Linear Regression (Mac & Windows)

Build a model from menus, read R² and coefficients, and check the diagnostics.

Comparisons

Stratum vs JMP

Where JMP's depth is worth it, and where a lighter native Mac & Windows app wins.

Statistics Explained

Understanding PCA Loadings & Scores

Loadings tell you what a component means; scores tell you where each observation sits.

How-To

How to Run PCA (Mac & Windows)

Reduce many variables to a few components; read the scree plot and biplot.

Case Studies

The Penguin Paradox: How to Tell Three Species Apart When the Average Lies

344 real penguins, a bill correlation that runs backwards until you split by species, then a decision tree, random forest and PCA that classify three species at ~98% — no code.

Case Studies

The Gap That Closed: 142 Countries, 55 Years, One Chart

Turn the most famous scatter in statistics into 142 connected trajectories and watch the development gap close — a full Gapminder analysis, no code, no animation.

Case Studies

The Star Wars Map: A Whole Survey in One Chart

1,186 people, 14 characters, one chart. Correspondence Analysis maps a whole survey — and Jar Jar sits alone in the corner. No code.

Case Studies

The Diamond Paradox: Why “Better” Diamonds Cost Less

53,940 real diamonds, three quality grades that all run backwards, and the charts, ANOVA, regression, PCA and decision tree that crack it — no code.

Case Studies

Does the Drug Work? It Depends Whom You Ask

2,139 patients in the ACTG 175 HIV trial: the overall average said the drug worked, one arm said it harmed, and the honest answer lived in the subgroups — paired t-test, effect size, responder rate, ANOVA, MANOVA and ANCOVA. No code.

Case Studies

In Control AND Capable: The Number That Passes a Failing Process

A capability report says Ppk 1.36 — a pass. On 200 real piston-ring measurements, SPC control charts catch the drift that single number hides. The Six-Sigma rule: control before capability. No code.

Statistics Explained

What Is a Control Chart? SPC & the Nelson Rules

How SPC tells ordinary variation from a real change — and what the Nelson rules detect.

How-To

How to Build a Control Chart (Mac & Windows)

I-MR and X-bar charts with control limits and automatic Nelson-rule flagging.

Comparisons

Best Statistics Software for Mac & Windows (2026)

Native apps vs Excel vs R/Python vs the big suites — an honest map of the options.

Statistics Explained

Correlation vs Causation: Pearson vs Spearman

What correlation does and doesn't mean, and when to use Pearson vs Spearman.

How-To

How to Make a Scatter Plot (Mac & Windows)

Plot two variables, add a line of best fit, and read the correlation.

Statistics Explained

Mean vs Median: Which Should You Report?

Two measures of center that agree for symmetric data and diverge for skewed data.

Product Updates

What's New in Stratum 1.2

A guided first-run tour, on-screen Hints, panel-by-panel control-chart zoom, pixel-exact 16:9 / 4K chart export, and cleaner below-zero bar labels.

Product Updates

What's New in Stratum 1.1

Two new Pro analyses, Connected Scatter, custom categorical ordering, skewness & kurtosis, and drag-to-export cards.

How-To

How to Make a Histogram (Mac & Windows)

See the shape of a variable — pick bins, read skew, and export.

Comparisons

Stratum vs Excel for Statistical Analysis

Where Excel runs out of road for real statistics — and what a native Mac & Windows app does differently.

Statistics Explained

ANOVA Assumptions Explained

Normality, equal variance, independence — what each assumption means and how to check it.

How-To

How to Run a One-Way ANOVA (Mac & Windows)

Compare three or more group means, read F and p, and run a Tukey post-hoc.

How-To

How to Run a t-Test (Mac & Windows) — Without R or Python

Compare two groups and read the result in under a minute — no code, no notebooks.

Statistics Explained

Box Plots Explained: Quartiles, IQR & Outliers

Median, quartiles, the IQR box, whiskers and outliers — read a box plot at a glance.

How-To

How to Make a Box Plot (Mac & Windows)

Compare groups and spot outliers with a box-and-whisker plot — in under a minute.

Product Updates

Introducing Stratum: Serious Statistics, Native to Mac & Windows

What Stratum is, who it's for, and why we built a native statistics app for Mac & Windows.

More coming
Every week

New posts twice a week

Case studies, explainers, and how-tos are on the way. Subscribe via RSS or follow the video series to keep up.