Description:
Abstract When Martin Bland was diagnosed with type II diabetes, he did what any medical statistician would do: he analysed his blood test results to find out whether alcohol consumption was associated with an increase or decrease in blood sugar levels