Upload a CSV, pick your columns, and get the missing-data patterns, the co-missingness map, how many rows listwise deletion would cost you, Little's MCAR test, and the same estimates recomputed four ways so you can see how much the handling rule is doing to your result. Free.
Free analyses run on up to 10,000 rows. Larger files are randomly sampled to that size — sign up to analyze your full dataset.
Auditing missingness patterns and testing MCAR...
Sent to — per-column missing rates, pattern analysis, co-missingness map, complete-case retention, Little's MCAR test, MAR evidence tests, imputation impact, R code, and AI insights.
Analyze another fileEvery cell of the mapped columns is classified as present or missing, and everything else is derived from that indicator matrix: per-column rates, the distinct patterns of joint missingness, a pairwise co-missingness map with an observed-versus-independent ratio, and the number of rows surviving listwise deletion. Little's MCAR test is computed directly — a maximum-likelihood mean and covariance are fitted across all the data by EM, then each pattern's observed means are compared against them and the pooled Mahalanobis discrepancy is referred to a chi-square distribution. Missingness on each column is then tested against the values of every other column (Welch t-test for numeric, chi-square for categorical, Holm-adjusted for multiplicity). Finally the same means and correlations are recomputed under complete-case deletion, available-case, mean imputation and deterministic regression imputation, with the spread between them and the share of each column's spread that each rule retains.
Use it when a dataset has gaps and you need to know whether they are bad enough to change your conclusion — before running a regression, a comparison, or anything that will silently drop incomplete rows.
Not a general dataset profiler — use the data profile tool for types, distributions and quality flags across every column. Not a normality or outlier screen — use the data screening tool for that. And not an imputation engine: it measures how much the handling rule moves your answer, it does not hand back a filled dataset.
Built for: Analysts, researchers and data scientists who need to defend how they handled the gaps in a dataset
Typical data source: Any spreadsheet or CSV with blanks in it — survey responses with skipped questions, records with fields that were never filled in, joined tables with unmatched rows
Any table with gaps in it, one record per row. For example, a patient records extract:
Minimum 20 rows · Best with 200-10,000 rows and 3-15 columns
Standard-library analysis: how bad is your missing data, and does it bias your results? Map the columns you care about and get the per-column missing rates, the pattern analysis (which columns go missing together, which is where the mechanism shows itself), a co-missingness map, how many rows survive listwise deletion, Little's MCAR test computed directly, per-column tests of whether missingness on one column is associated with the values of another — and then the impact half: the same means and correlations recomputed under complete-case deletion, available-case, mean imputation, and regression imputation, so you can see how far the answer moves when you change the handling rule.
How much of each column is missing — the first number everyone asks for, and the one that most understates the problem.
Which columns go missing together, and how much of the data each combination accounts for. This is where a mechanism shows itself.
A pairwise map of which columns fail in the same rows, compared against what independent missingness would produce.
Little's test of the hypothesis that values went missing completely at random, with the sample size and the assumptions behind the verdict.
Whether the rows missing a value differ from the rows that have it, on everything else you recorded — the practical evidence for MAR.
The same means and correlations under four handling rules, so the size of the choice is a number rather than an opinion.
How much of each column's spread survives each rule — the concrete cost of mean imputation, measured on your data.
Plain-English interpretation — what the numbers mean, what's significant, and what to do next.
How much of my data would a complete-case analysis throw away?
Map the columns your analysis uses. The audit reports the per-column rates, the distinct patterns of joint missingness, and the exact number of rows that survive listwise deletion — usually a much smaller number than any single column's missing rate suggests, because different rows are missing different fields.
See our FAQ for details on pricing, data privacy, and how the analysis works. Every report includes a Methodology section showing the statistical test, assumptions checked, and diagnostics run.
Run any analysis on your own data — validated R analyses, interactive reports, AI insights, and PDF export.
Try Free — No Credit CardTell us what went wrong, in your own words. We capture the page you're on automatically, so no need to describe where you are.