GP Lab Tech

Parametric two-sample comparison. Includes Shapiro-Wilk normality check.

Parametric comparison of 2+ groups. Includes Tukey HSD post-hoc.

Non-parametric alternative to t-test.

Non-parametric alternative to ANOVA.

Kaplan-Meier curves with log-rank test.

Binary outcome regression with odds ratios.

Pearson & Spearman correlation. Correlation does NOT imply causation.

Linear regression with validity checks.

Cronbach's alpha, split-half reliability, item-total correlations.

Chi-Square test of independence for categorical variables.

Chi-square goodness-of-fit test — chisq.test(x = observed, p = probs). Tests whether observed counts of a categorical column match a hypothesized distribution. Leave probabilities blank for a uniform null (equal probability per category).

Paired t-test for comparing two related measurements.

Two-way ANOVA with interaction effects.

Analysis of covariance — ANOVA with a continuous covariate.

Multivariate ANOVA for multiple dependent variables.

Non-parametric paired comparison (alternative to paired t-test).

Non-parametric repeated measures test (alternative to repeated ANOVA).

McNemar's test for paired binary data (before/after).

ROC curve and AUC for diagnostic/classification performance.

Bland-Altman analysis for method agreement.

Sensitivity, specificity, PPV, NPV, LR+, LR− for diagnostic tests.

Number Needed to Treat/Harm with risk differences.

Dose-response curve fitting (4PL or log-linear).

Principal Component Analysis for dimensionality reduction.

Exploratory Factor Analysis with varimax rotation.

Cluster analysis (K-Means or Hierarchical).

Stepwise regression with forward/backward/both selection.

Cohen's Kappa for inter-rater agreement (categorical).

Intraclass Correlation Coefficient for rater reliability.

Shapiro-Wilk, Anderson-Darling, Lilliefors, Jarque-Bera normality tests.

Statistical power and required sample size calculation.

Meta-analysis (DerSimonian-Laird) with fixed and random effects.

Density ridges showing distribution of a numeric value across categories. Top 20 categories by frequency are shown.

Publication-quality correlation heatmap, hierarchically clustered so similar variables sit together.

Visualises set intersections across binary columns. Accepts 0/1, TRUE/FALSE, Yes/No.

Circular flow diagram between two categorical columns. Top 50 flows are shown. Weight column is optional.

Interactive zoomable map with one marker per row. Requires latitude and longitude columns.

Interactive stacked time-series showing category shares over time. Date column can be ISO date or year.

ggplot2 scatter (geom_point) of two numeric columns. Optional categorical column to color points by group. Auto-detects Palmer Penguins species palette.

Bar chart with dplyr-pipe aggregation. Leave Value blank for category counts; pick a numeric Value + aggregation for grouped summaries. Optional Fill column for stacked / dodged bars.

Histogram of a numeric column with adjustable bins. Optional Fill By splits the histogram by category — "Overlay" position is best for comparing distributions.

Smooth kernel density estimate (geom_density) and / or frequency polygon (geom_freqpoly). Best for comparing distribution shapes across groups without binning artefacts.

Box-and-whisker plot — median, Q1/Q3, whiskers (1.5 × IQR), and outliers. Optional jitter overlay shows the actual data points alongside the summary.

Scatter matrix for 3-8 numeric columns. Lower triangle: pairwise scatter (optionally colored). Diagonal: density. Upper triangle: Pearson r with strength-shaded background. Best way to see all relationships at once.