Summary statistics, frequency tables, and visualisations for individual variables in your dataset.
Univariate analysis examines each variable in your dataset independently. For numeric variables Statulator produces a summary statistics table (mean, standard deviation, quartiles, minimum, maximum) together with a histogram, Q-Q plot, and box plot. For categorical variables it produces a frequency table (counts and percentages) together with a bar chart and pie chart.
When the data are loaded, Statulator automatically classifies each column as Numeric, Categorical, or ID using simple heuristics: columns with fewer than 6 unique values are treated as categorical; columns where every value is unique are flagged as identifiers (IDs). You can override these defaults in the variable-selection dialog before running the analysis.
A hospital quality team has exported a CSV file containing 250 patient records with columns for Age (years), LOS (length of stay, days), Gender (Male/Female), and Department (Emergency, Cardiology, Orthopaedics, General). They want a quick descriptive overview of every variable.
1 Open the Dataset Analysis page and click the file-load button to load your CSV. Your data stays in your browser.
2 Click Select Variables. Statulator lists all columns with auto-detected types. Verify that Age and LOS are Numeric, and Gender and Department are Categorical. Adjust if needed, then click Save Changes.
3 Click the UniVariate button.
4 Statulator generates an accordion panel for every selected variable. Expand any panel to see the statistics table and charts.
For Age (numeric): a table showing Mean = 54.2, SD = 16.8, Min = 18, Q1 = 42, Median = 55, Q3 = 67, Max = 93, n = 247, Missing = 3; a histogram of the age distribution; a Q-Q plot for normality assessment; and a box plot highlighting the quartiles and any outliers.
For Department (categorical): a frequency table listing each department with its count and percentage (e.g., Emergency 82, 32.8 %); a bar chart of frequencies; and a pie chart of the proportional breakdown.
The mean and standard deviation describe the centre and spread of numeric data. Compare the mean to the median: if they are close the distribution is roughly symmetric; a large gap suggests skewness. Q1 and Q3 define the interquartile range (IQR = Q3 − Q1), the middle 50 % of values.
The histogram shows the frequency distribution. Look for overall shape (symmetric, left-skewed, right-skewed), the number of modes (unimodal, bimodal), and any gaps or outlying bars far from the main mass.
The quantile-quantile plot compares your data to a theoretical normal distribution. Points lying close to the diagonal reference line indicate approximate normality. Systematic deviations at the tails (S-shape or banana curve) reveal skewness or heavy tails.
The box spans Q1 to Q3 with the median line inside. Whiskers extend to the most extreme data points within 1.5 × IQR of the box. Individual dots beyond the whiskers are potential outliers that warrant investigation.
The frequency table shows how observations are distributed across categories. Check for highly imbalanced groups (e.g., one category containing > 80 % of the data) which may affect downstream analyses. The bar chart makes magnitudes easy to compare; the pie chart highlights relative proportions.
Statulator uses the sample (Bessel-corrected) standard deviation with denominator \(n-1\).
Sort the \(n\) observations in ascending order. Let \(L_p = (n-1) \cdot p\) for the \(p\)-th percentile.
\[ Q_p = x_{\lfloor L_p \rfloor} + (L_p - \lfloor L_p \rfloor)(x_{\lfloor L_p \rfloor + 1} - x_{\lfloor L_p \rfloor}) \]quantile() default, NumPy’s numpy.quantile default, and Excel’s PERCENTILE.INC.Outlier fences: lower = \(Q_1 - 1.5\,\text{IQR}\), upper = \(Q_3 + 1.5\,\text{IQR}\).
Emergency department waiting times (minutes) for 150 patients: mean = 42, median = 35, SD = 22, IQR = 25–52, skewness = 1.4.
Analysis: The right skew (mean > median) and skewness of 1.4 indicate a long tail of very long waits. The histogram shows a peak near 30 min with a right tail. Report the median and IQR as the primary summary, since the distribution is not symmetric.
Final exam scores (0–100) for 200 students: mean = 72.4, median = 74, SD = 11.3, min = 28, max = 98.
Analysis: Mean and median are close, suggesting approximate symmetry. The Q-Q plot is roughly linear. The box plot shows two low outliers below 35. These students may need targeted support.
Annual household income ($) for 500 respondents: mean = $68,200, median = $52,000, SD = $45,000, IQR = $32k–$78k.
Analysis: The large gap between mean and median signals strong positive skew, typical of income distributions. A histogram would show a long right tail. The median ($52k) is the better central-tendency measure here.
Battery life (hours) for 60 smartphones: mean = 11.2, median = 11.0, SD = 1.8, range = 7.1–15.6.
Analysis: Near-symmetric (mean ≈ median). The Q-Q plot and Shapiro-Francia test (p = 0.34) support normality. Reporting mean ± SD (11.2 ± 1.8 hours) is appropriate.
Categorical: crop disease status across 300 plants. Healthy = 210 (70%), Mild = 55 (18.3%), Severe = 35 (11.7%).
Analysis: A frequency table and bar chart show that the majority of plants are healthy. A pie chart illustrates the proportions. The modal category is "Healthy."