Question 5: What would be the effect of the sampling method on the standard deviation of capillary glucose? How would the correlation with laboratory glucose be affected by the sampling method?
The sampling method drops everyone below 5.5 mmol/L and so removes observations at the lower extreme of the distribution. This reduces the spread of the capillary glucose measurements and so will reduce the standard deviation.
Reducing the variation in glucose should reduce the correlation coefficient.
Back to question.
This page maintained by Martin Bland.
Last updated: 1 March, 2007.