Statistical Analysis Dashboard - v.2

📊 Statistical Analysis Dashboard v2 — Interactive HTML File
By Data'sOk
━━━━━━━━━━━━━━━━━━━━━━━━
A complete statistical toolkit. Drop in a CSV, get real analysis — not just a pretty chart.
This is the second generation of the very first tool Data'sOk ever built. The original was an Excel dashboard for descriptive stats, correlation, and regression. This version keeps everything that worked and adds the things Excel was never going to do well: an interactive correlation heatmap, proper outlier detection and one-click cleanup, automatic forecasting with method comparison, and a full hypothesis-testing suite — all in a single file that opens in any browser.
No macros. No add-ins. No subscription. Upload a CSV, and you have a working analysis in under a minute.
━━━━━━━━━━━━━━━━━━━━━━━━
WHAT YOU GET
7 analysis tabs, all driven by your own data:
📁 Data
Drag-and-drop CSV upload, or paste/type data directly into a spreadsheet-style grid. Year/date columns are detected automatically and excluded from statistics. Sortable columns. Built-in outlier detection (IQR, Z-score, or Modified Z-score) with one-click replacement via linear interpolation.
📊 Descriptive Statistics
19 measures per variable — mean, geometric mean, harmonic mean, quartiles, mode, standard deviation, variance, coefficient of variation, skewness, kurtosis, growth rates, and more. Sortable table, variable comparison chart, and a CV ranking chart to spot your most volatile variables instantly.
📈 Distribution
Histogram with adjustable bins, a properly rendered box plot with outliers marked, a sequence/time plot, and an automatic shape analysis (skewness, kurtosis, normality signal) in plain language.
🔗 Correlation
Click any cell on the interactive heatmap to instantly plot that pair's scatter diagram. Trend line, R², p-value, and strength rating included. Top-correlations panel surfaces your strongest relationships automatically.
📉 Regression
Full multiple OLS regression with a clickable variable picker. Coefficients, standard errors, t-stats, p-values, R², adjusted R², residual plots, and — the part people actually use — a plain-English interpretation of what the model is telling you.
🔮 Forecast
Four forecasting methods (Naïve, Moving Average, Exponential Smoothing, Linear Trend) compared automatically on a hold-out sample. Best method selected for you, with confidence intervals and a full forecast table. Hover any method for a one-line explanation of how it works.
🧪 Hypothesis Tests
One-sample and two-sample t-tests, a normality test, and an F-test for variance equality — each with a plain pass/fail readout, not just a p-value to interpret yourself.
━━━━━━━━━━━━━━━━━━━━━━━━
ADDITIONAL FEATURES
🎨 6 light themes and 4 dark themes, or set your own colours
🖊️ Logo upload and editable title for white-labelling
💾 Export data or full statistics to CSV, copy a KPI summary, or print to PDF
📖 Built-in manual covering every tab and statistical method
✅ 3 sample datasets (economics, sales, health) to explore before using your own data
━━━━━━━━━━━━━━━━━━━━━━━━
WHO IT'S FOR
→ Analysts who want a fast first look at a new dataset without opening R or Python
→ Researchers who need descriptive stats, correlation, and regression without a stats package licence
→ Consultants who want a polished, presentable analysis to hand to a client
→ Students and self-learners who want to see statistical concepts (skewness, p-values, R²) explained alongside the numbers
→ Anyone who has a CSV and a question, and wants an answer in minutes, not hours
━━━━━━━━━━━━━━━━━━━━━━━━
REQUIREMENTS
Any modern browser — Chrome, Firefox, Safari, or Edge. Works on Windows, Mac, and Linux. No installation, no internet connection required after download. Your data stays on your machine — nothing is uploaded anywhere.