| This post was kindly contributed by The SAS Training Post - go there to comment and to read the full post. |
As far as numbers go, the number zero is rather mysterious for data. Is it something or is it nothing?
What happens when you have missing data but enter 0? 
This topic triggered an intriguing discussion in my recent Programming 2: Data Manipulation Techniques class. In this post I’d like to compare various industries to show you their universal take on the 0.
To start off, see if you agree with the analysts in different industries who entered zero in place of missing values below:
Can you have a 0 for blood glucose levels?
A health analyst is missing blood glucose level records for several patients and decides to enter 0 as a value. But health policy makers know that a 0 blood glucose level is highly unlikely as the patient would go into insulin shock and need immediate medical attention.
Can you have a 0% interest rate?
A customer is currently negotiating interest rates and nothing is confirmed yet. Meanwhile a mortgage specialist enters 0% with the intention to come back later and change it to a valid value. But he goes away on extended leave. What we have now is a very happy customer who scans her mortgage statements with glee thanking her stars for the day she met this banker. From the bank’s perspective however, banks might need to declare bankruptcy if they started doling money out without charging any interest.
Can you have a 0 degree Celsius temperature recording in Toronto on July 10th?
A weather specialist doesn’t have records for several cities on July 10th. She goes ahead and enters 0 for all the missing values. What she’s effectively done by entering a 0 value is to possibly mislead the general public. Just picture all the summer Toronto visitors who take this at face value and come prepared with warm clothing for a cold July!
Continue reading “Is something missing or is it a ZERO?”
| This post was kindly contributed by The SAS Training Post - go there to comment and to read the full post. |