An instrument maker has randomly selected and submitted 20 of its electronic blood-pressure measurement units to a medical laboratory for testing. At the lab, each unit measures the "blood pressure" of a testing device that simulates a human arm; within the device the pressure is a known constant. For the instruments submitted for testing, the standard deviation of the readings is 1.4 millimeters.
Assuming a normal distribution of measurement values, determine the 95% confidence interval for the population standard deviation of instruments produced by this manufacturer.