Evidence-appraisal glossary
Hosmer-Lemeshow Test
The Hosmer-Lemeshow test is a classic check of whether a risk model's predictions match observed outcomes across groups of patients sorted by predicted risk. A large p-value is read as reassurance of good calibration.
It sorts patients into bins, often ten, by predicted risk and compares predicted with observed event counts in each. The test has well known weaknesses: it depends on the arbitrary number of bins, has little power in small samples and too much in large ones, and says nothing about the direction or size of any miscalibration. Many methodologists now prefer calibration plots and the calibration slope, which show where and how a model goes wrong.
Read the full Reading the Evidence blog.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.