I have a big dataset of sensor data (body temperature) acquired on a lot of people and each person is assessed for several days.

person1 = [temp_day1, temp_day2, .... , temp_dayn]

During each day the temperature is assessed 24 times (1 every hour).

temp_day1 = [temp_01h, temp_02h, ....,temp_24h]

Sensors could have some problems...for example they could be not calibrated and they could be influenced by enviromental temperature etc..

In order to process the data (compare temperature across the people) I had in mind to standardize the data of each subject according to the mean and standard deviation of his temperature during all his assessed days.

Is that correct? Do dividing for the standard deviation have any meaning? Are there other ways to make the data comparable?




Browse other questions tagged or ask your own question.