Would you happen to have a source for this? I'm finding it hard to imagine how a study like this can even take place given that there have (probably) not been many survivors among those who have received a dose of 4 Sieverts.
This is called the "linear hypothesis" and is widely used. They use it in medicine to decide if diagnostic scans are worth the risk. The risk could actually be lower; some scientists think there's a threshold below which there's no risk. There's too much statistical uncertainty at low exposure levels to know for sure.
This is the course site, click webcast lectures on the side: http://muller.lbl.gov/teaching/physics10/pffp.html