Interpreting results with GLI-2012 using Z-score and LLN
Let’s try to calculate %Predicted and Z-Score
The heights of the dogs (in mm) are: 600, 470, 170, 430 and 300
CLASSIC METHOD: %PREDICTED = MEASURED / PREDICTED x 100
Rottweiler 600/394 x 100 = 152% taller then PREDICTED
Dachshund 170/394 x 100 = 43% shorter then PREDICTED
Schnauzer 430/394 x 100 = 109% very close to PREDICTED
It is a very intuitive method and indicates what is the percentage value of the MEASURED VALUE in relation to the PREDICTED VALUE
NEW METHOD: Z–SCORE = (MEASURED – PREDICTED) / STANDARD DEVIATION
Rottweiler (600-394)/147 = +1.40= more then 1 SD taller
Dachshund (170-394)/147 = -1.52= more then 1 SD shorter
Schnauzer (430-394)/147 = +0.24= less then 1 SD taller
This method is less intuitive and indicates how often a measurement is greater (+) or less (-) of the STANDARD DEVIATION.