Raw Gyro data-type

[Jay]

I dumped the raw gyro data value based on the http://yujinrobot.github.io/kobuki/doxygen/enAppendixProtocolSpecification.html section "Raw Data Of 3D Gyro".

For this study the turtlebot was stationary. Shown below is part of the output :

19, ffd8, ffc4

12, ffe8, ffbf

4, fff6, ffbb

fffd, fffd, ffbd

3, fffb, ffc8

10, fff5, ffd5

1d, fff0, ffd9

24, ffec, ffd1

I was hoping that the data is just "raw" concatenated msb and lsb bytes from the STM L3G4200D. 

But it looks like the output is a 2's complement number instead of "RAW" gyro readings. Can anyone confirm if indeed this is the case ? The comment says the data is : "ADC output of each-axis in 0.00875 deg/s". Is this correct ?