Enter Dataset Values
Paste rows, upload a CSV file, or load the example dataset.
Worked Example Dataset
| StudyHours | SleepHours | Attendance | ExamScore |
|---|---|---|---|
| 2 | 8 | 75 | 60 |
| 3 | 7 | 80 | 65 |
| 4 | 7 | 85 | 72 |
| 5 | 6 | 88 | 78 |
| 6 | 6 | 90 | 82 |
| 7 | 5 | 92 | 86 |
| 8 | 5 | 95 | 91 |
This sample helps demonstrate positive, negative, and moderate relationships across several variables.
Core Statistical Formulas
1. Mean
mean = Σx / n
2. Z Score
z = (x - mean) / standard deviation
3. Covariance
cov(X,Y) = Σ[(xi - meanX)(yi - meanY)] / (n-1) for sample mode, or divide by n for population mode.
4. Pearson Correlation
r = cov(X,Y) / (sdX × sdY)
Z scores standardize each variable onto the same scale. The correlation matrix then compares standardized relationships without the original unit sizes distorting the comparison.
Step by Step
- Enter a dataset name for organized exports.
- Paste rows of numeric data or upload a CSV file.
- Choose the correct delimiter or leave auto detect enabled.
- Check the first-row-header option when column names are included.
- Select sample or population standard deviation mode.
- Choose the number of decimal places to display.
- Press the calculation button to generate the statistics.
- Review the z scores, covariance matrix, correlation matrix, and heatmap.
- Download CSV files or export the full result section as PDF.
FAQs
1. What does this calculator produce?
It calculates descriptive statistics, z scores for each observation, a covariance matrix, a Pearson correlation matrix, and a heatmap for visual pattern checking.
2. Why are z scores useful here?
Z scores standardize variables with different units. That makes comparisons easier and helps you see which observations are above or below each variable's mean.
3. Can I upload a file instead of pasting data?
Yes. Upload a CSV or text file, choose the delimiter, and the page will parse the numeric values directly.
4. What happens if one variable never changes?
Its standard deviation becomes zero. Z scores for that variable are set to zero, and related correlations are shown as N/A because they are undefined.
5. Should I choose sample or population mode?
Use sample mode when your data represents a subset from a larger group. Use population mode when your dataset includes the entire group you want to describe.
6. Can this handle tab-separated values?
Yes. You can paste tab-separated rows, choose tab manually, or leave auto detect enabled.
7. Why are some correlations close to zero?
A value near zero means the variables do not show a strong linear relationship. They may still have a weak, curved, or irregular pattern.
8. Does correlation prove causation?
No. Correlation only measures association. It cannot confirm whether one variable directly causes changes in another.