Estimating accuracy when the reference standard is continuous
You have been given a datasheet "validity2.csv". This sheet is part of the the corrected data of a study done in the blood bank AIIMS, Rishikesh. The study The study was about the evaluation of two separate point-of-care devices to estimate hemoglobin, which would be useful in donor selection in blood donation camps. The hemoglobin could be tested both on capillary blood after a finger prick or on venous samples. Using a finger prick is more convenient, but we had to check whether using finger prick blood samples for hemoglobin would be accurate enough. Also we decided to check the accuracy of the instruments on venous blood, as a fall back if capillary hemoglobin was not accurate enough. Before routine use, the instruments needed to be validated. In the original study, we checked the accuracy of capillary and venous hemoglobin estimation on two different instruments of two different instrument models (total of four different instruments). We have included here the ve