Recovery at two hospitals — summary table

The aggregated 2 × 2 × 2 table behind hospital_outcomes: for each fictional hospital and severity, the number of patients, the number recovered and the recovery rate. Four rows that are enough to discover Simpson's paradox by hand.

Open in the lab Download CSV

Rows: 4
Columns: 5
One row is: one hospital × severity group
Version: v1

Data dictionary

Column	Type	Unit	Summary	Description
`hospital` Hospital	category		2 levels: Hospital A, Hospital B	Treating hospital (fictional).
`severity` Severity	ordinal		2 levels: mild, severe	Severity on admission.
`patients` Patients	integer	patients	mean 250, median 250, SD 173.2, range 100–400	Number of patients.
`recovered` Recovered	integer	patients	mean 206.3, median 197.5, SD 145.3, range 70–360	Number who recovered.
`recovery_rate_pct` Recovery rate	number	%	mean 82.5, median 82.5, SD 11.9, range 70–95	recovered ÷ patients × 100.

What it is good for

Combine strata to get overall rates and see the reversal.
Compute weighted averages of stratum rates with different weights.

Methods: proportions, weighted average, stratification, direct standardisation.

Data quality notes

Aggregated data: 4 rows representing 1 000 patients.
No missing values.

Provenance and licence

Exact counts fixed for teaching; they equal the row counts in hospital_outcomes.

Synthetic data · CC0-1.0 (synthetic data generated by Statistics Quest) · generated deterministically (seed 20241021). It does not describe real people or organisations; any resemblance is coincidental.

Preview

Dataset

Recovery at two hospitals — summary table

4 rows · one row = one hospital × severity group · synthetic data

About this dataset Open in the lab

Hospital	Severity	Patients (patients)	Recovered (patients)	Recovery rate (%)
Hospital A	mild	100	95	95
Hospital A	severe	400	300	75
Hospital B	mild	400	360	90
Hospital B	severe	100	70	70

Used in

Confounding and causation · Statistical Thinking
Simpson's paradox · Resampling and Statistical Reasoning
Rates and proportions in public health · Applied Statistics for Professionals
Causal inference: potential outcomes and causal diagrams · Advanced Foundations