Topic hub
Data Analysis
Practical workflows for exploring, cleaning, and interpreting real datasets, from the first look at the data through to a conclusion you can defend.
Box Plot Interpretation: IQR and Outliers
Box plot interpretation made practical: the five-number summary, IQR, the 1.5×IQR outlier rule, sample-size guidance, and comparing groups.
Box-Cox Transformation: When and How to Use It
This Box-Cox transformation guide covers the lambda formula, picking lambda by profile likelihood, a worked income example, and honest reporting tips.
Choose a Correlation Test: Pearson vs Spearman
Choose a correlation test with a quick checklist: Pearson, Spearman, or Kendall, assumption checks, a worked example, and alternative methods.
Cohen's d Calculation and Interpretation
Compute Cohen's d with step by step formulas, worked examples, and small sample correction (Hedges' g). Convert d to U3, probability, or NNT.
Communicating Uncertainty: A Practitioner Guide
A practitioner's guide to communicating uncertainty: pair a point estimate with a labeled range, name assumptions, and cite PNAS and WHO evidence.
Conditional Probability Exercises With Solutions
A hands-on guide to conditional probability exercises: worked formulas, tree diagrams, joint probability tables, and three solved practice problems.
Confidence Interval vs Prediction Interval
Confidence interval vs prediction interval, compared: a CI estimates a population parameter, while a PI bounds one future value and is always wider.
Confidence Level vs Significance Level Explained
Confidence level vs significance level explained: significance level sets your false-positive risk in a test, and confidence level is its complement.
Data Visualization Best Practices Guide
These data visualization best practices cover five perception backed rules for chart types, encoding comparisons, and color, to avoid misleading charts.
Dual-Axis Charts: When to Use Them Safely
Dual-axis charts pair two scales on one plot, but they can mislead when ranges are not synchronized; here is when they are justified and how to check them.
Exploratory Data Analysis: A Practical Workflow
The exploratory data analysis workflow analysts follow: data quality checks, missingness patterns, distribution shape, outliers, and when to stop.
Heteroscedasticity Test: Which One to Use
Run a heteroscedasticity test by plotting residuals, then confirm with Breusch-Pagan, Koenker, or White, and switch to HC3 robust standard errors if it fails.
Histogram Interpretation: A Practitioner Guide
Histogram interpretation means checking six things in order: shape, center, spread, skew, outliers, and modes, plus a worked example and calculator links.
Histogram vs Boxplot: When to Use Each
Histogram vs boxplot: use the 30-observation rule to pick the right chart, know which settings to record, and see when combining both works best.
Log Transform Data: When and How to Do It
Log transform data when values span multiple orders of magnitude: how to decide, apply it in Excel, R, or Python, and back-transform the results.
Misleading Graphs: How to Spot Them
Misleading graphs can make readers misjudge trends 6 to 15 times more often. Use this 30-second SCAM checklist to spot axis and scale tricks fast.
Missing Data Imputation: A Five-Step Workflow
A five-step workflow for missing data imputation: decide when to impute or drop values, match methods to inference or prediction, then test sensitivity.
Multiple Regression Diagnostics Guide
A practitioner's guide to multiple regression diagnostics: run the four assumption checks, test multicollinearity with VIF, and avoid common reporting mistakes.
Nonparametric Tests: When and How to Use Them
Nonparametric tests compare groups or measure association without assuming a normal distribution; use this map to match your data to the right test.
Normality Tests: Choosing the Right Test
A practical guide to choosing and interpreting normality tests, with rules of thumb by sample size and what to do when data fail the test.
Odds Ratio Interpretation: A Worked Guide
Odds ratio interpretation in plain English: odds versus probability, a worked 6x example, and a checklist for reporting an OR in your own research.
One-Tailed vs Two-Tailed Tests: How to Choose
A practical checklist for choosing one-tailed vs two-tailed tests: four questions, worked examples, and why switching tails after seeing data is invalid.
Pearson vs Spearman: When to Use Each One
Compare Pearson vs Spearman correlation with a plot-first checklist, worked examples, and guidance on choosing the right coefficient for your data.
Scatter Plot Interpretation Guide
A practical guide to scatter plot interpretation: a 6-step checklist covering form, direction, strength, and outliers before trusting a correlation number.
T-Test vs Z-Test: How to Choose
Choose a t-test vs z-test by whether the population standard deviation is known, not sample size. Includes a 6-step checklist and a worked example.